fiftyone.core.dataset¶

FiftyOne datasets.

Copyright 2017-2024, Voxel51, Inc.
voxel51.com

Exceptions:

DatasetNotFoundError(name)

Exception raised when a dataset is not found.

Functions:

`list_datasets`([glob_patt, tags, info])	Lists the available FiftyOne datasets.
`dataset_exists`(name)	Checks if the dataset exists.
`load_dataset`(name[, create_if_necessary])	Loads the FiftyOne dataset with the given name.
`get_default_dataset_name`()	Returns a default dataset name based on the current time.
`make_unique_dataset_name`(root)	Makes a unique dataset name with the given root name.
`get_default_dataset_dir`(name)	Returns the default dataset directory for the dataset with the given name.
`delete_dataset`(name[, verbose])	Deletes the FiftyOne dataset with the given name.
`delete_datasets`(glob_patt[, verbose])	Deletes all FiftyOne datasets whose names match the given glob pattern.
`delete_non_persistent_datasets`([verbose])	Deletes all non-persistent datasets.

Classes:

Dataset([name, _create])

A FiftyOne dataset.

exception fiftyone.core.dataset.DatasetNotFoundError(name)¶

Bases: ValueError

Exception raised when a dataset is not found.

args¶

with_traceback()¶: Exception.with_traceback(tb) – set self.__traceback__ to tb and return self.

fiftyone.core.dataset.list_datasets(glob_patt=None, tags=None, info=False)¶

Lists the available FiftyOne datasets.

Parameters

glob_patt (None) – an optional glob pattern of names to return
tags (None) – only include datasets that have the specified tag or list of tags
info (False) – whether to return info dicts describing each dataset rather than just their names

Returns

a list of dataset names or info dicts

fiftyone.core.dataset.dataset_exists(name)¶

Checks if the dataset exists.

Parameters: name – the name of the dataset
Returns: True/False

fiftyone.core.dataset.load_dataset(name, create_if_necessary=False)¶

Loads the FiftyOne dataset with the given name.

To create a new dataset, use the Dataset constructor.

Note

Dataset instances are singletons keyed by their name, so all calls to this method with a given dataset name in a program will return the same object.

Parameters

name – the name of the dataset
create_if_necessary (False) – if no dataset exists, create an empty one

Raises

DatasetNotFoundError – if the dataset does not exist and create_if_necessary is False

Returns

a Dataset

fiftyone.core.dataset.get_default_dataset_name()¶

Returns a default dataset name based on the current time.

Returns: a dataset name

fiftyone.core.dataset.make_unique_dataset_name(root)¶

Makes a unique dataset name with the given root name.

Parameters: root – the root name for the dataset
Returns: the dataset name

fiftyone.core.dataset.get_default_dataset_dir(name)¶

Returns the default dataset directory for the dataset with the given name.

Parameters: name – the dataset name
Returns: the default directory for the dataset

fiftyone.core.dataset.delete_dataset(name, verbose=False)¶

Deletes the FiftyOne dataset with the given name.

Parameters

name – the name of the dataset
verbose (False) – whether to log the name of the deleted dataset

fiftyone.core.dataset.delete_datasets(glob_patt, verbose=False)¶

Deletes all FiftyOne datasets whose names match the given glob pattern.

Parameters

glob_patt – a glob pattern of datasets to delete
verbose (False) – whether to log the names of deleted datasets

fiftyone.core.dataset.delete_non_persistent_datasets(verbose=False)¶

Deletes all non-persistent datasets.

Parameters: verbose (False) – whether to log the names of deleted datasets

class fiftyone.core.dataset.Dataset(name=None, _create=True, *args, **kwargs)¶

Bases: fiftyone.core.collections.SampleCollection

A FiftyOne dataset.

Datasets represent an ordered collection of fiftyone.core.sample.Sample instances that describe a particular type of raw media (e.g., images or videos) together with a user-defined set of fields.

FiftyOne datasets ingest and store the labels for all samples internally; raw media is stored on disk and the dataset provides paths to the data.

See this page for an overview of working with FiftyOne datasets.

Parameters

name (None) – the name of the dataset. By default, get_default_dataset_name() is used
persistent (False) – whether the dataset should persist in the database after the session terminates
overwrite (False) – whether to overwrite an existing dataset of the same name

Attributes:

`media_type`	The media type of the dataset.
`group_field`	The group field of the dataset, or None if the dataset is not grouped.
`group_slice`	The current group slice of the dataset, or None if the dataset is not grouped.
`group_slices`	The list of group slices of the dataset, or None if the dataset is not grouped.
`group_media_types`	A dict mapping group slices to media types, or None if the dataset is not grouped.
`default_group_slice`	The default group slice of the dataset, or None if the dataset is not grouped.
`version`	The version of the `fiftyone` package for which the dataset is formatted.
`name`	The name of the dataset.
`slug`	The slug of the dataset.
`created_at`	The datetime that the dataset was created.
`last_modified_at`	The datetime that the dataset was last modified.
`last_loaded_at`	The datetime that the dataset was last loaded.
`persistent`	Whether the dataset persists in the database after a session is terminated.
`tags`	A list of tags on the dataset.
`description`	A string description on the dataset.
`info`	A user-facing dictionary of information about the dataset.
`app_config`	A `fiftyone.core.odm.dataset.DatasetAppConfig` that customizes how this dataset is visualized in the FiftyOne App.
`classes`	A dict mapping field names to list of class label strings for the corresponding fields of the dataset.
`default_classes`	A list of class label strings for all `fiftyone.core.labels.Label` fields of this dataset that do not have customized classes defined in `classes()`.
`mask_targets`	A dict mapping field names to mask target dicts, each of which defines a mapping between pixel values (2D masks) or RGB hex strings (3D masks) and label strings for the segmentation masks in the corresponding field of the dataset.
`default_mask_targets`	A dict defining a default mapping between pixel values (2D masks) or RGB hex strings (3D masks) and label strings for the segmentation masks of all `fiftyone.core.labels.Segmentation` fields of this dataset that do not have customized mask targets defined in `mask_targets()`.
`skeletons`	A dict mapping field names to `fiftyone.core.odm.dataset.KeypointSkeleton` instances, each of which defines the semantic labels and point connectivity for the `fiftyone.core.labels.Keypoint` instances in the corresponding field of the dataset.
`default_skeleton`	A default `fiftyone.core.odm.dataset.KeypointSkeleton` defining the semantic labels and point connectivity for all `fiftyone.core.labels.Keypoint` fields of this dataset that do not have customized skeletons defined in `skeleton()`.
`deleted`	Whether the dataset is deleted.
`has_saved_views`	Whether this dataset has any saved views.
`has_workspaces`	Whether this dataset has any saved workspaces.
`has_annotation_runs`	Whether this collection has any annotation runs.
`has_brain_runs`	Whether this collection has any brain runs.
`has_evaluations`	Whether this collection has any evaluation results.
`has_runs`	Whether this collection has any runs.

Methods:

`summary`()	Returns a string summary of the dataset.
`stats`([include_media, include_indexes, …])	Returns stats about the dataset on disk.
`first`()	Returns the first sample in the dataset.
`last`()	Returns the last sample in the dataset.
`head`([num_samples])	Returns a list of the first few samples in the dataset.
`tail`([num_samples])	Returns a list of the last few samples in the dataset.
`one`(expr[, exact])	Returns a single sample in this dataset matching the expression.
`view`()	Returns a `fiftyone.core.view.DatasetView` containing the entire dataset.
`get_field_schema`([ftype, embedded_doc_type, …])	Returns a schema dictionary describing the fields of the samples in the dataset.
`get_frame_field_schema`([ftype, …])	Returns a schema dictionary describing the fields of the frames of the samples in the dataset.
`add_sample_field`(field_name, ftype[, …])	Adds a new sample field or embedded field to the dataset, if necessary.
`add_dynamic_sample_fields`([fields, …])	Adds all dynamic sample fields to the dataset’s schema.
`add_frame_field`(field_name, ftype[, …])	Adds a new frame-level field or embedded field to the dataset, if necessary.
`list_summary_fields`()	Lists the summary fields on the dataset.
`create_summary_field`(path[, field_name, …])	Populates a sample-level field that records the unique values or numeric ranges that appear in the specified field on each sample in the dataset.
`check_summary_fields`()	Returns a list of summary fields that may need to be updated.
`update_summary_field`(field_name)	Updates the summary field based on the current values of its source field.
`delete_summary_field`(field_name[, error_level])	Deletes the summary field from all samples in the dataset.
`delete_summary_fields`(field_names[, error_level])	Deletes the summary fields from all samples in the dataset.
`add_dynamic_frame_fields`([fields, …])	Adds all dynamic frame fields to the dataset’s schema.
`add_group_field`(field_name[, default, …])	Adds a group field to the dataset, if necessary.
`rename_sample_field`(field_name, new_field_name)	Renames the sample field to the given new name.
`rename_sample_fields`(field_mapping)	Renames the sample fields to the given new names.
`rename_frame_field`(field_name, new_field_name)	Renames the frame-level field to the given new name.
`rename_frame_fields`(field_mapping)	Renames the frame-level fields to the given new names.
`clone_sample_field`(field_name, new_field_name)	Clones the given sample field into a new field of the dataset.
`clone_sample_fields`(field_mapping)	Clones the given sample fields into new fields of the dataset.
`clone_frame_field`(field_name, new_field_name)	Clones the frame-level field into a new field.
`clone_frame_fields`(field_mapping)	Clones the frame-level fields into new fields.
`clear_sample_field`(field_name)	Clears the values of the field from all samples in the dataset.
`clear_sample_fields`(field_names)	Clears the values of the fields from all samples in the dataset.
`clear_frame_field`(field_name)	Clears the values of the frame-level field from all samples in the dataset.
`clear_frame_fields`(field_names)	Clears the values of the frame-level fields from all samples in the dataset.
`delete_sample_field`(field_name[, error_level])	Deletes the field from all samples in the dataset.
`delete_sample_fields`(field_names[, error_level])	Deletes the fields from all samples in the dataset.
`remove_dynamic_sample_field`(field_name[, …])	Removes the dynamic embedded sample field from the dataset’s schema.
`remove_dynamic_sample_fields`(field_names[, …])	Removes the dynamic embedded sample fields from the dataset’s schema.
`delete_frame_field`(field_name[, error_level])	Deletes the frame-level field from all samples in the dataset.
`delete_frame_fields`(field_names[, error_level])	Deletes the frame-level fields from all samples in the dataset.
`remove_dynamic_frame_field`(field_name[, …])	Removes the dynamic embedded frame field from the dataset’s schema.
`remove_dynamic_frame_fields`(field_names[, …])	Removes the dynamic embedded frame fields from the dataset’s schema.
`add_group_slice`(name, media_type)	Adds a group slice with the given media type to the dataset, if necessary.
`rename_group_slice`(name, new_name)	Renames the group slice with the given name.
`delete_group_slice`(name)	Deletes all samples in the given group slice from the dataset.
`iter_samples`([progress, autosave, …])	Returns an iterator over the samples in the dataset.
`iter_groups`([group_slices, progress, …])	Returns an iterator over the groups in the dataset.
`get_group`(group_id[, group_slices])	Returns a dict containing the samples for the given group ID.
`add_sample`(sample[, expand_schema, dynamic, …])	Adds the given sample to the dataset.
`add_samples`(samples[, expand_schema, …])	Adds the given samples to the dataset.
`add_collection`(sample_collection[, …])	Adds the contents of the given collection to the dataset.
`merge_sample`(sample[, key_field, …])	Merges the fields of the given sample into this dataset.
`merge_samples`(samples[, key_field, key_fcn, …])	Merges the given samples into this dataset.
`delete_samples`(samples_or_ids)	Deletes the given sample(s) from the dataset.
`delete_frames`(frames_or_ids)	Deletes the given frames(s) from the dataset.
`delete_groups`(groups_or_ids)	Deletes the given groups(s) from the dataset.
`delete_labels`([labels, ids, tags, view, fields])	Deletes the specified labels from the dataset.
`save`()	Saves the dataset to the database.
`has_saved_view`(name)	Whether this dataset has a saved view with the given name.
`list_saved_views`([info])	List saved views on this dataset.
`save_view`(name, view[, description, color, …])	Saves the given view into this dataset under the given name so it can be loaded later via `load_saved_view()`.
`get_saved_view_info`(name)	Loads the editable information about the saved view with the given name.
`update_saved_view_info`(name, info)	Updates the editable information for the saved view with the given name.
`load_saved_view`(name)	Loads the saved view with the given name.
`delete_saved_view`(name)	Deletes the saved view with the given name.
`delete_saved_views`()	Deletes all saved views from this dataset.
`has_workspace`(name)	Whether this dataset has a saved workspace with the given name.
`list_workspaces`([info])	List saved workspaces on this dataset.
`save_workspace`(name, workspace[, …])	Saves a workspace into this dataset under the given name so it can be loaded later via `load_workspace()`.
`load_workspace`(name)	Loads the saved workspace with the given name.
`get_workspace_info`(name)	Gets the information about the workspace with the given name.
`update_workspace_info`(name, info)	Updates the editable information for the saved view with the given name.
`delete_workspace`(name)	Deletes the saved workspace with the given name.
`delete_workspaces`()	Deletes all saved workspaces from this dataset.
`clone`([name, persistent])	Creates a copy of the dataset.
`clear`()	Removes all samples from the dataset.
`clear_frames`()	Removes all frame labels from the dataset.
`ensure_frames`()	Ensures that the video dataset contains frame instances for every frame of each sample’s source video.
`delete`()	Deletes the dataset.
`add_dir`([dataset_dir, dataset_type, …])	Adds the contents of the given directory to the dataset.
`merge_dir`([dataset_dir, dataset_type, …])	Merges the contents of the given directory into the dataset.
`add_archive`(archive_path[, dataset_type, …])	Adds the contents of the given archive to the dataset.
`merge_archive`(archive_path[, dataset_type, …])	Merges the contents of the given archive into the dataset.
`add_importer`(dataset_importer[, …])	Adds the samples from the given `fiftyone.utils.data.importers.DatasetImporter` to the dataset.
`merge_importer`(dataset_importer[, …])	Merges the samples from the given `fiftyone.utils.data.importers.DatasetImporter` into the dataset.
`add_images`(paths_or_samples[, …])	Adds the given images to the dataset.
`add_labeled_images`(samples, sample_parser[, …])	Adds the given labeled images to the dataset.
`add_images_dir`(images_dir[, tags, …])	Adds the given directory of images to the dataset.
`add_images_patt`(images_patt[, tags, progress])	Adds the given glob pattern of images to the dataset.
`ingest_images`(paths_or_samples[, …])	Ingests the given iterable of images into the dataset.
`ingest_labeled_images`(samples, sample_parser)	Ingests the given iterable of labeled image samples into the dataset.
`add_videos`(paths_or_samples[, …])	Adds the given videos to the dataset.
`add_labeled_videos`(samples, sample_parser[, …])	Adds the given labeled videos to the dataset.
`add_videos_dir`(videos_dir[, tags, …])	Adds the given directory of videos to the dataset.
`add_videos_patt`(videos_patt[, tags, progress])	Adds the given glob pattern of videos to the dataset.
`ingest_videos`(paths_or_samples[, …])	Ingests the given iterable of videos into the dataset.
`ingest_labeled_videos`(samples, sample_parser)	Ingests the given iterable of labeled video samples into the dataset.
`from_dir`([dataset_dir, dataset_type, …])	Creates a `Dataset` from the contents of the given directory.
`from_archive`(archive_path[, dataset_type, …])	Creates a `Dataset` from the contents of the given archive.
`from_importer`(dataset_importer[, name, …])	Creates a `Dataset` by importing the samples in the given `fiftyone.utils.data.importers.DatasetImporter`.
`from_images`(paths_or_samples[, …])	Creates a `Dataset` from the given images.
`from_labeled_images`(samples, sample_parser)	Creates a `Dataset` from the given labeled images.
`from_images_dir`(images_dir[, name, …])	Creates a `Dataset` from the given directory of images.
`from_images_patt`(images_patt[, name, …])	Creates a `Dataset` from the given glob pattern of images.
`from_videos`(paths_or_samples[, …])	Creates a `Dataset` from the given videos.
`from_labeled_videos`(samples, sample_parser)	Creates a `Dataset` from the given labeled videos.
`from_videos_dir`(videos_dir[, name, …])	Creates a `Dataset` from the given directory of videos.
`from_videos_patt`(videos_patt[, name, …])	Creates a `Dataset` from the given glob pattern of videos.
`from_dict`(d[, name, persistent, overwrite, …])	Loads a `Dataset` from a JSON dictionary generated by `fiftyone.core.collections.SampleCollection.to_dict()`.
`from_json`(path_or_str[, name, persistent, …])	Loads a `Dataset` from JSON generated by `fiftyone.core.collections.SampleCollection.write_json()` or `fiftyone.core.collections.SampleCollection.to_json()`.
`reload`()	Reloads the dataset and any in-memory samples from the database.
`clear_cache`()	Clears the dataset’s in-memory cache.
`add_stage`(stage)	Applies the given `fiftyone.core.stages.ViewStage` to the collection.
`aggregate`(aggregations)	Aggregates one or more `fiftyone.core.aggregations.Aggregation` instances.
`annotate`(anno_key[, label_schema, …])	Exports the samples and optional label field(s) in this collection to the given annotation backend.
`apply_model`(model[, label_field, …])	Applies the model to the samples in the collection.
`bounds`(field_or_expr[, expr, safe])	Computes the bounds of a numeric field of the collection.
`compute_embeddings`(model[, …])	Computes embeddings for the samples in the collection using the given model.
`compute_metadata`([overwrite, num_workers, …])	Populates the `metadata` field of all samples in the collection.
`compute_patch_embeddings`(model, patches_field)	Computes embeddings for the image patches defined by `patches_field` of the samples in the collection using the given model.
`concat`(samples)	Concatenates the contents of the given `SampleCollection` to this collection.
`count`([field_or_expr, expr, safe])	Counts the number of field values in the collection.
`count_label_tags`([label_fields])	Counts the occurrences of all label tags in the specified label field(s) of this collection.
`count_sample_tags`()	Counts the occurrences of sample tags in this collection.
`count_values`(field_or_expr[, expr, safe])	Counts the occurrences of field values in the collection.
`create_index`(field_or_spec[, unique, wait])	Creates an index on the given field or with the given specification, if necessary.
`delete_annotation_run`(anno_key)	Deletes the annotation run with the given key from this collection.
`delete_annotation_runs`()	Deletes all annotation runs from this collection.
`delete_brain_run`(brain_key)	Deletes the brain method run with the given key from this collection.
`delete_brain_runs`()	Deletes all brain method runs from this collection.
`delete_evaluation`(eval_key)	Deletes the evaluation results associated with the given evaluation key from this collection.
`delete_evaluations`()	Deletes all evaluation results from this collection.
`delete_run`(run_key)	Deletes the run with the given key from this collection.
`delete_runs`()	Deletes all runs from this collection.
`distinct`(field_or_expr[, expr, safe])	Computes the distinct values of a field in the collection.
`draw_labels`(output_dir[, rel_dir, …])	Renders annotated versions of the media in the collection with the specified label data overlaid to the given directory.
`drop_index`(field_or_name)	Drops the index for the given field or name, if necessary.
`evaluate_classifications`(pred_field[, …])	Evaluates the classification predictions in this collection with respect to the specified ground truth labels.
`evaluate_detections`(pred_field[, gt_field, …])	Evaluates the specified predicted detections in this collection with respect to the specified ground truth detections.
`evaluate_regressions`(pred_field[, gt_field, …])	Evaluates the regression predictions in this collection with respect to the specified ground truth values.
`evaluate_segmentations`(pred_field[, …])	Evaluates the specified semantic segmentation masks in this collection with respect to the specified ground truth masks.
`exclude`(sample_ids)	Excludes the samples with the given IDs from the collection.
`exclude_by`(field, values)	Excludes the samples with the given field values from the collection.
`exclude_fields`([field_names, meta_filter, …])	Excludes the fields with the given names from the samples in the collection.
`exclude_frames`(frame_ids[, omit_empty])	Excludes the frames with the given IDs from the video collection.
`exclude_groups`(group_ids)	Excludes the groups with the given IDs from the grouped collection.
`exclude_labels`([labels, ids, tags, fields, …])	Excludes the specified labels from the collection.
`exists`(field[, bool])	Returns a view containing the samples in the collection that have (or do not have) a non-`None` value for the given field or embedded field.
`export`([export_dir, dataset_type, …])	Exports the samples in the collection to disk.
`filter_field`(field, filter[, only_matches])	Filters the values of a field or embedded field of each sample in the collection.
`filter_keypoints`(field[, filter, labels, …])	Filters the individual `fiftyone.core.labels.Keypoint.points` elements in the specified keypoints field of each sample in the collection.
`filter_labels`(field, filter[, only_matches, …])	Filters the `fiftyone.core.labels.Label` field of each sample in the collection.
`flatten`([stages])	Returns a flattened view that contains all samples in the dynamic grouped collection.
`geo_near`(point[, location_field, …])	Sorts the samples in the collection by their proximity to a specified geolocation.
`geo_within`(boundary[, location_field, strict])	Filters the samples in this collection to only include samples whose geolocation is within a specified boundary.
`get_annotation_info`(anno_key)	Returns information about the annotation run with the given key on this collection.
`get_brain_info`(brain_key)	Returns information about the brain method run with the given key on this collection.
`get_classes`(field)	Gets the classes list for the given field, or None if no classes are available.
`get_dynamic_field_schema`([fields, recursive])	Returns a schema dictionary describing the dynamic fields of the samples in the collection.
`get_dynamic_frame_field_schema`([fields, …])	Returns a schema dictionary describing the dynamic fields of the frames in the collection.
`get_evaluation_info`(eval_key)	Returns information about the evaluation with the given key on this collection.
`get_field`(path[, ftype, embedded_doc_type, …])	Returns the field instance of the provided path, or `None` if one does not exist.
`get_index_information`([include_stats])	Returns a dictionary of information about the indexes on this collection.
`get_mask_targets`(field)	Gets the mask targets for the given field, or None if no mask targets are available.
`get_run_info`(run_key)	Returns information about the run with the given key on this collection.
`get_skeleton`(field)	Gets the keypoint skeleton for the given field, or None if no skeleton is available.
`group_by`(field_or_expr[, order_by, reverse, …])	Creates a view that groups the samples in the collection by a specified field or expression.
`has_annotation_run`(anno_key)	Whether this collection has an annotation run with the given key.
`has_brain_run`(brain_key)	Whether this collection has a brain method run with the given key.
`has_classes`(field)	Determines whether this collection has a classes list for the given field.
`has_evaluation`(eval_key)	Whether this collection has an evaluation with the given key.
`has_field`(path)	Determines whether the collection has a field with the given name.
`has_frame_field`(path)	Determines whether the collection has a frame-level field with the given name.
`has_mask_targets`(field)	Determines whether this collection has mask targets for the given field.
`has_run`(run_key)	Whether this collection has a run with the given key.
`has_sample_field`(path)	Determines whether the collection has a sample field with the given name.
`has_skeleton`(field)	Determines whether this collection has a keypoint skeleton for the given field.
`histogram_values`(field_or_expr[, expr, …])	Computes a histogram of the field values in the collection.
`init_run`(**kwargs)	Initializes a config instance for a new run.
`init_run_results`(run_key, **kwargs)	Initializes a results instance for the run with the given key.
`limit`(limit)	Returns a view with at most the given number of samples.
`limit_labels`(field, limit)	Limits the number of `fiftyone.core.labels.Label` instances in the specified labels list field of each sample in the collection.
`list_aggregations`()	Returns a list of all available methods on this collection that apply `fiftyone.core.aggregations.Aggregation` operations to this collection.
`list_annotation_runs`([type, method])	Returns a list of annotation keys on this collection.
`list_brain_runs`([type, method])	Returns a list of brain keys on this collection.
`list_evaluations`([type, method])	Returns a list of evaluation keys on this collection.
`list_indexes`()	Returns the list of index names on this collection.
`list_runs`(**kwargs)	Returns a list of run keys on this collection.
`list_schema`(field_or_expr[, expr])	Extracts the value type(s) in a specified list field across all samples in the collection.
`list_view_stages`()	Returns a list of all available methods on this collection that apply `fiftyone.core.stages.ViewStage` operations to this collection.
`load_annotation_results`(anno_key[, cache])	Loads the results for the annotation run with the given key on this collection.
`load_annotation_view`(anno_key[, select_fields])	Loads the `fiftyone.core.view.DatasetView` on which the specified annotation run was performed on this collection.
`load_annotations`(anno_key[, dest_field, …])	Downloads the labels from the given annotation run from the annotation backend and merges them into this collection.
`load_brain_results`(brain_key[, cache, load_view])	Loads the results for the brain method run with the given key on this collection.
`load_brain_view`(brain_key[, select_fields])	Loads the `fiftyone.core.view.DatasetView` on which the specified brain method run was performed on this collection.
`load_evaluation_results`(eval_key[, cache])	Loads the results for the evaluation with the given key on this collection.
`load_evaluation_view`(eval_key[, select_fields])	Loads the `fiftyone.core.view.DatasetView` on which the specified evaluation was performed on this collection.
`load_run_results`(run_key[, cache, load_view])	Loads the results for the run with the given key on this collection.
`load_run_view`(run_key[, select_fields])	Loads the `fiftyone.core.view.DatasetView` on which the specified run was performed on this collection.
`make_unique_field_name`([root])	Makes a unique field name with the given root name for the collection.
`map_labels`(field, map)	Maps the `label` values of a `fiftyone.core.labels.Label` field to new values for each sample in the collection.
`match`(filter)	Filters the samples in the collection by the given filter.
`match_frames`(filter[, omit_empty])	Filters the frames in the video collection by the given filter.
`match_labels`([labels, ids, tags, filter, …])	Selects the samples from the collection that contain (or do not contain) at least one label that matches the specified criteria.
`match_tags`(tags[, bool, all])	Returns a view containing the samples in the collection that have or don’t have any/all of the given tag(s).
`mean`(field_or_expr[, expr, safe])	Computes the arithmetic mean of the field values of the collection.
`merge_labels`(in_field, out_field)	Merges the labels from the given input field into the given output field of the collection.
`mongo`(pipeline[, _needs_frames, _group_slices])	Adds a view stage defined by a raw MongoDB aggregation pipeline.
`quantiles`(field_or_expr, quantiles[, expr, safe])	Computes the quantile(s) of the field values of a collection.
`register_run`(run_key, config[, results, …])	Registers a run under the given key on this collection.
`rename_annotation_run`(anno_key, new_anno_key)	Replaces the key for the given annotation run with a new key.
`rename_brain_run`(brain_key, new_brain_key)	Replaces the key for the given brain run with a new key.
`rename_evaluation`(eval_key, new_eval_key)	Replaces the key for the given evaluation with a new key.
`rename_run`(run_key, new_run_key)	Replaces the key for the given run with a new key.
`save_context`([batch_size, batching_strategy])	Returns a context that can be used to save samples from this collection according to a configurable batching strategy.
`save_run_results`(run_key, results[, …])	Saves run results for the run with the given key.
`schema`(field_or_expr[, expr, dynamic_only, …])	Extracts the names and types of the attributes of a specified embedded document field across all samples in the collection.
`select`(sample_ids[, ordered])	Selects the samples with the given IDs from the collection.
`select_by`(field, values[, ordered])	Selects the samples with the given field values from the collection.
`select_fields`([field_names, meta_filter, …])	Selects only the fields with the given names from the samples in the collection.
`select_frames`(frame_ids[, omit_empty])	Selects the frames with the given IDs from the video collection.
`select_group_slices`([slices, media_type, …])	Selects the samples in the group collection from the given slice(s).
`select_groups`(group_ids[, ordered])	Selects the groups with the given IDs from the grouped collection.
`select_labels`([labels, ids, tags, fields, …])	Selects only the specified labels from the collection.
`set_field`(field, expr[, _allow_missing])	Sets a field or embedded field on each sample in a collection by evaluating the given expression.
`set_label_values`(field_name, values[, …])	Sets the fields of the specified labels in the collection to the given values.
`set_values`(field_name, values[, key_field, …])	Sets the field or embedded field on each sample or frame in the collection to the given values.
`shuffle`([seed])	Randomly shuffles the samples in the collection.
`skip`(skip)	Omits the given number of samples from the head of the collection.
`sort_by`(field_or_expr[, reverse, create_index])	Sorts the samples in the collection by the given field(s) or expression(s).
`sort_by_similarity`(query[, k, reverse, …])	Sorts the collection by similarity to a specified query.
`split_labels`(in_field, out_field[, filter])	Splits the labels from the given input field into the given output field of the collection.
`std`(field_or_expr[, expr, safe, sample])	Computes the standard deviation of the field values of the collection.
`sum`(field_or_expr[, expr, safe])	Computes the sum of the field values of the collection.
`sync_last_modified_at`([include_frames])	Syncs the `last_modified_at` property(s) of the dataset.
`tag_labels`(tags[, label_fields])	Adds the tag(s) to all labels in the specified label field(s) of this collection, if necessary.
`tag_samples`(tags)	Adds the tag(s) to all samples in this collection, if necessary.
`take`(size[, seed])	Randomly samples the given number of samples from the collection.
`to_clips`(field_or_expr, **kwargs)	Creates a view that contains one sample per clip defined by the given field or expression in the video collection.
`to_dict`([rel_dir, include_private, …])	Returns a JSON dictionary representation of the collection.
`to_evaluation_patches`(eval_key, **kwargs)	Creates a view based on the results of the evaluation with the given key that contains one sample for each true positive, false positive, and false negative example in the collection, respectively.
`to_frames`(**kwargs)	Creates a view that contains one sample per frame in the video collection.
`to_json`([rel_dir, include_private, …])	Returns a JSON string representation of the collection.
`to_patches`(field, **kwargs)	Creates a view that contains one sample per object patch in the specified field of the collection.
`to_trajectories`(field, **kwargs)	Creates a view that contains one clip for each unique object trajectory defined by their `(label, index)` in a frame-level field of a video collection.
`untag_labels`(tags[, label_fields])	Removes the tag from all labels in the specified label field(s) of this collection, if necessary.
`untag_samples`(tags)	Removes the tag(s) from all samples in this collection, if necessary.
`update_run_config`(run_key, config)	Updates the run config for the run with the given key.
`validate_field_type`(path[, ftype, …])	Validates that the collection has a field of the given type.
`validate_fields_exist`(fields[, include_private])	Validates that the collection has field(s) with the given name(s).
`values`(field_or_expr[, expr, missing_value, …])	Extracts the values of a field from all samples in the collection.
`write_json`(json_path[, rel_dir, …])	Writes the colllection to disk in JSON format.

property media_type¶: The media type of the dataset.

property group_field¶

The group field of the dataset, or None if the dataset is not grouped.

Examples:

import fiftyone as fo
import fiftyone.zoo as foz

dataset = foz.load_zoo_dataset("quickstart-groups")

print(dataset.group_field)
# group

property group_slice¶

The current group slice of the dataset, or None if the dataset is not grouped.

Examples:

import fiftyone as fo
import fiftyone.zoo as foz

dataset = foz.load_zoo_dataset("quickstart-groups")

print(dataset.group_slices)
# ['left', 'right', 'pcd']

print(dataset.group_slice)
# left

# Change the current group slice
dataset.group_slice = "right"

print(dataset.group_slice)
# right

property group_slices¶

The list of group slices of the dataset, or None if the dataset is not grouped.

Examples:

import fiftyone as fo
import fiftyone.zoo as foz

dataset = foz.load_zoo_dataset("quickstart-groups")

print(dataset.group_slices)
# ['left', 'right', 'pcd']

property group_media_types¶

A dict mapping group slices to media types, or None if the dataset is not grouped.

Examples:

import fiftyone as fo
import fiftyone.zoo as foz

dataset = foz.load_zoo_dataset("quickstart-groups")

print(dataset.group_media_types)
# {'left': 'image', 'right': 'image', 'pcd': 'point-cloud'}

property default_group_slice¶

The default group slice of the dataset, or None if the dataset is not grouped.

Examples:

import fiftyone as fo
import fiftyone.zoo as foz

dataset = foz.load_zoo_dataset("quickstart-groups")

print(dataset.default_group_slice)
# left

# Change the default group slice
dataset.default_group_slice = "right"

print(dataset.default_group_slice)
# right

property version¶: The version of the fiftyone package for which the dataset is formatted.

property name¶: The name of the dataset.

property slug¶: The slug of the dataset.

property created_at¶: The datetime that the dataset was created.

property last_modified_at¶: The datetime that the dataset was last modified.

property last_loaded_at¶: The datetime that the dataset was last loaded.

property persistent¶: Whether the dataset persists in the database after a session is terminated.

property tags¶

A list of tags on the dataset.

Examples:

import fiftyone as fo

dataset = fo.Dataset()

# Add some tags
dataset.tags = ["test", "projectA"]

# Edit the tags
dataset.tags.pop()
dataset.tags.append("projectB")
dataset.save()  # must save after edits

property description¶

A string description on the dataset.

Examples:

import fiftyone as fo

dataset = fo.Dataset()

# Store a description on the dataset
dataset.description = "Your description here"

property info¶

A user-facing dictionary of information about the dataset.

Examples:

import fiftyone as fo

dataset = fo.Dataset()

# Store a class list in the dataset's info
dataset.info = {"classes": ["cat", "dog"]}

# Edit the info
dataset.info["other_classes"] = ["bird", "plane"]
dataset.save()  # must save after edits

property app_config¶

A fiftyone.core.odm.dataset.DatasetAppConfig that customizes how this dataset is visualized in the FiftyOne App.

Examples:

import fiftyone as fo
import fiftyone.utils.image as foui
import fiftyone.zoo as foz

dataset = foz.load_zoo_dataset("quickstart")

# View the dataset's current App config
print(dataset.app_config)

# Generate some thumbnail images
foui.transform_images(
    dataset,
    size=(-1, 32),
    output_field="thumbnail_path",
    output_dir="/tmp/thumbnails",
)

# Modify the dataset's App config
dataset.app_config.media_fields = ["filepath", "thumbnail_path"]
dataset.app_config.grid_media_field = "thumbnail_path"
dataset.save()  # must save after edits

session = fo.launch_app(dataset)

property classes¶

A dict mapping field names to list of class label strings for the corresponding fields of the dataset.

Examples:

import fiftyone as fo

dataset = fo.Dataset()

# Set classes for the `ground_truth` and `predictions` fields
dataset.classes = {
    "ground_truth": ["cat", "dog"],
    "predictions": ["cat", "dog", "other"],
}

# Edit an existing classes list
dataset.classes["ground_truth"].append("other")
dataset.save()  # must save after edits

property default_classes¶

A list of class label strings for all fiftyone.core.labels.Label fields of this dataset that do not have customized classes defined in classes().

Examples:

import fiftyone as fo

dataset = fo.Dataset()

# Set default classes
dataset.default_classes = ["cat", "dog"]

# Edit the default classes
dataset.default_classes.append("rabbit")
dataset.save()  # must save after edits

property mask_targets¶

A dict mapping field names to mask target dicts, each of which defines a mapping between pixel values (2D masks) or RGB hex strings (3D masks) and label strings for the segmentation masks in the corresponding field of the dataset.

Examples:

import fiftyone as fo

#
# 2D masks
#

dataset = fo.Dataset()

# Set mask targets for the `ground_truth` and `predictions` fields
dataset.mask_targets = {
    "ground_truth": {1: "cat", 2: "dog"},
    "predictions": {1: "cat", 2: "dog", 255: "other"},
}

# Or, for RGB mask targets
dataset.mask_targets = {
    "segmentations": {"#3f0a44": "road", "#eeffee": "building", "#ffffff": "other"}
}

# Edit an existing mask target
dataset.mask_targets["ground_truth"][255] = "other"
dataset.save()  # must save after edits

#
# 3D masks
#

dataset = fo.Dataset()

# Set mask targets for the `ground_truth` and `predictions` fields
dataset.mask_targets = {
    "ground_truth": {"#499CEF": "cat", "#6D04FF": "dog"},
    "predictions": {
        "#499CEF": "cat", "#6D04FF": "dog", "#FF6D04": "person"
    },
}

# Edit an existing mask target
dataset.mask_targets["ground_truth"]["#FF6D04"] = "person"
dataset.save()  # must save after edits

property default_mask_targets¶

A dict defining a default mapping between pixel values (2D masks) or RGB hex strings (3D masks) and label strings for the segmentation masks of all fiftyone.core.labels.Segmentation fields of this dataset that do not have customized mask targets defined in mask_targets().

Examples:

import fiftyone as fo

#
# 2D masks
#

dataset = fo.Dataset()

# Set default mask targets
dataset.default_mask_targets = {1: "cat", 2: "dog"}

# Or, for RGB mask targets
dataset.default_mask_targets = {"#3f0a44": "road", "#eeffee": "building", "#ffffff": "other"}

# Edit the default mask targets
dataset.default_mask_targets[255] = "other"
dataset.save()  # must save after edits

#
# 3D masks
#

dataset = fo.Dataset()

# Set default mask targets
dataset.default_mask_targets = {"#499CEF": "cat", "#6D04FF": "dog"}

# Edit the default mask targets
dataset.default_mask_targets["#FF6D04"] = "person"
dataset.save()  # must save after edits

property skeletons¶

A dict mapping field names to fiftyone.core.odm.dataset.KeypointSkeleton instances, each of which defines the semantic labels and point connectivity for the fiftyone.core.labels.Keypoint instances in the corresponding field of the dataset.

Examples:

import fiftyone as fo

dataset = fo.Dataset()

# Set keypoint skeleton for the `ground_truth` field
dataset.skeletons = {
    "ground_truth": fo.KeypointSkeleton(
        labels=[
            "left hand" "left shoulder", "right shoulder", "right hand",
            "left eye", "right eye", "mouth",
        ],
        edges=[[0, 1, 2, 3], [4, 5, 6]],
    )
}

# Edit an existing skeleton
dataset.skeletons["ground_truth"].labels[-1] = "lips"
dataset.save()  # must save after edits

property default_skeleton¶

A default fiftyone.core.odm.dataset.KeypointSkeleton defining the semantic labels and point connectivity for all fiftyone.core.labels.Keypoint fields of this dataset that do not have customized skeletons defined in skeleton().

Examples:

import fiftyone as fo

dataset = fo.Dataset()

# Set default keypoint skeleton
dataset.default_skeleton = fo.KeypointSkeleton(
    labels=[
        "left hand" "left shoulder", "right shoulder", "right hand",
        "left eye", "right eye", "mouth",
    ],
    edges=[[0, 1, 2, 3], [4, 5, 6]],
)

# Edit the default skeleton
dataset.default_skeleton.labels[-1] = "lips"
dataset.save()  # must save after edits

property deleted¶: Whether the dataset is deleted.

summary()¶

Returns a string summary of the dataset.

Returns: a string summary

stats(include_media=False, include_indexes=False, compressed=False)¶

Returns stats about the dataset on disk.

The samples keys refer to the sample documents stored in the database.

For video datasets, the frames keys refer to the frame documents stored in the database.

The media keys refer to the raw media associated with each sample on disk.

The index[es] keys refer to the indexes associated with the dataset.

Note that dataset-level metadata such as annotation runs are not included in this computation.

Parameters

include_media (False) – whether to include stats about the size of the raw media in the dataset
include_indexes (False) – whether to include stats on the dataset’s indexes
compressed (False) – whether to return the sizes of collections in their compressed form on disk (True) or the logical uncompressed size of the collections (False)

Returns

a stats dict

first()¶

Returns the first sample in the dataset.

Returns: a fiftyone.core.sample.Sample

last()¶

Returns the last sample in the dataset.

Returns: a fiftyone.core.sample.Sample

head(num_samples=3)¶

Returns a list of the first few samples in the dataset.

If fewer than num_samples samples are in the dataset, only the available samples are returned.

Parameters: num_samples (3) – the number of samples
Returns: a list of fiftyone.core.sample.Sample objects

tail(num_samples=3)¶

Returns a list of the last few samples in the dataset.

If fewer than num_samples samples are in the dataset, only the available samples are returned.

Parameters: num_samples (3) – the number of samples
Returns: a list of fiftyone.core.sample.Sample objects

one(expr, exact=False)¶

Returns a single sample in this dataset matching the expression.

Examples:

import fiftyone as fo
import fiftyone.zoo as foz
from fiftyone import ViewField as F

dataset = foz.load_zoo_dataset("quickstart")

#
# Get a sample by filepath
#

# A random filepath in the dataset
filepath = dataset.take(1).first().filepath

# Get sample by filepath
sample = dataset.one(F("filepath") == filepath)

#
# Dealing with multiple matches
#

# Get a sample whose image is JPEG
sample = dataset.one(F("filepath").ends_with(".jpg"))

# Raises an error since there are multiple JPEGs
dataset.one(F("filepath").ends_with(".jpg"), exact=True)

Parameters

expr – a fiftyone.core.expressions.ViewExpression or MongoDB expression that evaluates to True for the sample to match
exact (False) – whether to raise an error if multiple samples match the expression

Returns

a fiftyone.core.sample.Sample

view()¶

Returns a fiftyone.core.view.DatasetView containing the entire dataset.

Returns: a fiftyone.core.view.DatasetView

get_field_schema(ftype=None, embedded_doc_type=None, read_only=None, info_keys=None, created_after=None, include_private=False, flat=False, mode=None)¶

Returns a schema dictionary describing the fields of the samples in the dataset.

Parameters

ftype (None) – an optional field type or iterable of types to which to restrict the returned schema. Must be subclass(es) of fiftyone.core.fields.Field
embedded_doc_type (None) – an optional embedded document type or iterable of types to which to restrict the returned schema. Must be subclass(es) of fiftyone.core.odm.BaseEmbeddedDocument
read_only (None) – whether to restrict to (True) or exclude (False) read-only fields. By default, all fields are included
info_keys (None) – an optional key or list of keys that must be in the field’s info dict
created_after (None) – an optional datetime specifying a minimum creation date
include_private (False) – whether to include fields that start with _ in the returned schema
flat (False) – whether to return a flattened schema where all embedded document fields are included as top-level keys
mode (None) – whether to apply the above constraints before and/or after flattening the schema. Only applicable when flat is True. Supported values are ("before", "after", "both"). The default is "after"

Returns

a dict mapping field names to fiftyone.core.fields.Field instances

get_frame_field_schema(ftype=None, embedded_doc_type=None, read_only=None, info_keys=None, created_after=None, include_private=False, flat=False, mode=None)¶

Returns a schema dictionary describing the fields of the frames of the samples in the dataset.

Only applicable for datasets that contain videos.

Parameters

ftype (None) – an optional field type or iterable of types to which to restrict the returned schema. Must be subclass(es) of fiftyone.core.fields.Field
embedded_doc_type (None) – an optional embedded document type or iterable of types to which to restrict the returned schema. Must be subclass(es) of fiftyone.core.odm.BaseEmbeddedDocument
read_only (None) – whether to restrict to (True) or exclude (False) read-only fields. By default, all fields are included
info_keys (None) – an optional key or list of keys that must be in the field’s info dict
created_after (None) – an optional datetime specifying a minimum creation date
include_private (False) – whether to include fields that start with _ in the returned schema
flat (False) – whether to return a flattened schema where all embedded document fields are included as top-level keys
mode (None) – whether to apply the above constraints before and/or after flattening the schema. Only applicable when flat is True. Supported values are ("before", "after", "both"). The default is "after"

Returns

a dict mapping field names to fiftyone.core.fields.Field instances, or None if the dataset does not contain videos

add_sample_field(field_name, ftype, embedded_doc_type=None, subfield=None, fields=None, description=None, info=None, read_only=False, **kwargs)¶

Adds a new sample field or embedded field to the dataset, if necessary.

Parameters

field_name – the field name or embedded.field.name
ftype – the field type to create. Must be a subclass of fiftyone.core.fields.Field
embedded_doc_type (None) – the fiftyone.core.odm.BaseEmbeddedDocument type of the field. Only applicable when ftype is fiftyone.core.fields.EmbeddedDocumentField
subfield (None) – the fiftyone.core.fields.Field type of the contained field. Only applicable when ftype is fiftyone.core.fields.ListField or fiftyone.core.fields.DictField
fields (None) – a list of fiftyone.core.fields.Field instances defining embedded document attributes. Only applicable when ftype is fiftyone.core.fields.EmbeddedDocumentField
description (None) – an optional description
info (None) – an optional info dict
read_only (False) – whether the field should be read-only

Raises

ValueError – if a field of the same name already exists and it is not compliant with the specified values

add_dynamic_sample_fields(fields=None, recursive=True, add_mixed=False)¶

Adds all dynamic sample fields to the dataset’s schema.

Dynamic fields are embedded document fields with at least one non-None value that have not been declared on the dataset’s schema.

Parameters

fields (None) – an optional field or iterable of fields for which to add dynamic fields. By default, all fields are considered
recursive (True) – whether to recursively inspect nested lists and embedded documents for dynamic fields
add_mixed (False) – whether to declare fields that contain values of mixed types as generic fiftyone.core.fields.Field instances (True) or to skip such fields (False)

add_frame_field(field_name, ftype, embedded_doc_type=None, subfield=None, fields=None, description=None, info=None, read_only=False, **kwargs)¶

Adds a new frame-level field or embedded field to the dataset, if necessary.

Only applicable to datasets that contain videos.

Parameters

field_name – the field name or embedded.field.name
ftype – the field type to create. Must be a subclass of fiftyone.core.fields.Field
embedded_doc_type (None) – the fiftyone.core.odm.BaseEmbeddedDocument type of the field. Only applicable when ftype is fiftyone.core.fields.EmbeddedDocumentField
subfield (None) – the fiftyone.core.fields.Field type of the contained field. Only applicable when ftype is fiftyone.core.fields.ListField or fiftyone.core.fields.DictField
fields (None) – a list of fiftyone.core.fields.Field instances defining embedded document attributes. Only applicable when ftype is fiftyone.core.fields.EmbeddedDocumentField
description (None) – an optional description
info (None) – an optional info dict
read_only (False) – whether the field should be read-only

Raises

ValueError – if a field of the same name already exists and it is not compliant with the specified values

list_summary_fields()¶

Lists the summary fields on the dataset.

Use create_summary_field() to create summary fields, and use delete_summary_field() to delete them.

Returns: a list of summary field names

create_summary_field(path, field_name=None, sidebar_group=None, include_counts=False, group_by=None, read_only=True, create_index=True)¶

Populates a sample-level field that records the unique values or numeric ranges that appear in the specified field on each sample in the dataset.

This method is particularly useful for summarizing frame-level fields of video datasets, in which case the sample-level field records the unique values or numeric ranges that appear in the specified frame-level field across all frames of that sample. This summary field can then be efficiently queried to retrieve samples that contain specific values of interest in at least one frame.

Examples:

import fiftyone as fo
import fiftyone.zoo as foz
from fiftyone import ViewField as F

dataset = foz.load_zoo_dataset("quickstart-video")
dataset.set_field("frames.detections.detections.confidence", F.rand()).save()

# Generate a summary field for object labels
dataset.create_summary_field("frames.detections.detections.label")

# Generate a summary field for [min, max] confidences
dataset.create_summary_field("frames.detections.detections.confidence")

# Generate a summary field for object labels and counts
dataset.create_summary_field(
    "frames.detections.detections.label",
    field_name="frames_detections_label2",
    include_counts=True,
)

# Generate a summary field for per-label [min, max] confidences
dataset.create_summary_field(
    "frames.detections.detections.confidence",
    field_name="frames_detections_confidence2",
    group_by="label",
)

print(dataset.list_summary_fields())

Parameters

path – an input field path
field_name (None) – the sample-level field in which to store the summary data. By default, a suitable name is derived from the given path
sidebar_group (None) – the name of a App sidebar group to which to add the summary field. By default, all summary fields are added to a "summaries" group. You can pass False to skip sidebar group modification
include_counts (False) – whether to include per-value counts when summarizing categorical fields
group_by (None) – an optional attribute to group by when path is a numeric field to generate per-attribute [min, max] ranges. This may either be an absolute path or an attribute name that is interpreted relative to path
read_only (True) – whether to mark the summary field as read-only
create_index (True) – whether to create database index(es) for the summary field

Returns

the summary field name

check_summary_fields()¶

Returns a list of summary fields that may need to be updated.

Summary fields may need to be updated whenever there have been modifications to the dataset’s samples since the summaries were last generated.

Note that inclusion in this list is only a heuristic, as any sample modifications may not have affected the summary’s source field.

Returns: list of summary field names

update_summary_field(field_name)¶

Updates the summary field based on the current values of its source field.

Parameters: field_name – the summary field

delete_summary_field(field_name, error_level=0)¶

Deletes the summary field from all samples in the dataset.

Parameters

field_name – the summary field
error_level (0) – the error level to use. Valid values are:
0 (-) – raise error if a summary field cannot be deleted
1 (-) – log warning if a summary field cannot be deleted
2 (-) – ignore summary fields that cannot be deleted

delete_summary_fields(field_names, error_level=0)¶

Deletes the summary fields from all samples in the dataset.

Parameters

field_names – the summary field or iterable of summary fields
error_level (0) – the error level to use. Valid values are:
0 (-) – raise error if a summary field cannot be deleted
1 (-) – log warning if a summary field cannot be deleted
2 (-) – ignore summary fields that cannot be deleted

add_dynamic_frame_fields(fields=None, recursive=True, add_mixed=False)¶

Adds all dynamic frame fields to the dataset’s schema.

Dynamic fields are embedded document fields with at least one non-None value that have not been declared on the dataset’s schema.

Parameters

fields (None) – an optional field or iterable of fields for which to add dynamic fields. By default, all fields are considered
recursive (True) – whether to recursively inspect nested lists and embedded documents for dynamic fields
add_mixed (False) – whether to declare fields that contain values of mixed types as generic fiftyone.core.fields.Field instances (True) or to skip such fields (False)

add_group_field(field_name, default=None, description=None, info=None, read_only=False)¶

Adds a group field to the dataset, if necessary.

Parameters

field_name – the field name
default (None) – a default group slice for the field
description (None) – an optional description
info (None) – an optional info dict
read_only (False) – whether the field should be read-only

Raises

ValueError – if a group field with another name already exists

rename_sample_field(field_name, new_field_name)¶

Renames the sample field to the given new name.

You can use dot notation (embedded.field.name) to rename embedded fields.

Parameters

field_name – the field name or embedded.field.name
new_field_name – the new field name or embedded.field.name

rename_sample_fields(field_mapping)¶

Renames the sample fields to the given new names.

You can use dot notation (embedded.field.name) to rename embedded fields.

Parameters: field_mapping – a dict mapping field names to new field names

rename_frame_field(field_name, new_field_name)¶

Renames the frame-level field to the given new name.

You can use dot notation (embedded.field.name) to rename embedded frame fields.

Only applicable to datasets that contain videos.

Parameters

field_name – the field name or embedded.field.name
new_field_name – the new field name or embedded.field.name

rename_frame_fields(field_mapping)¶

Renames the frame-level fields to the given new names.

You can use dot notation (embedded.field.name) to rename embedded frame fields.

Parameters: field_mapping – a dict mapping field names to new field names

clone_sample_field(field_name, new_field_name)¶

Clones the given sample field into a new field of the dataset.

You can use dot notation (embedded.field.name) to clone embedded fields.

Parameters

field_name – the field name or embedded.field.name
new_field_name – the new field name or embedded.field.name

clone_sample_fields(field_mapping)¶

Clones the given sample fields into new fields of the dataset.

You can use dot notation (embedded.field.name) to clone embedded fields.

Parameters: field_mapping – a dict mapping field names to new field names into which to clone each field

clone_frame_field(field_name, new_field_name)¶

Clones the frame-level field into a new field.

You can use dot notation (embedded.field.name) to clone embedded frame fields.

Only applicable to datasets that contain videos.

Parameters

field_name – the field name or embedded.field.name
new_field_name – the new field name or embedded.field.name

clone_frame_fields(field_mapping)¶

Clones the frame-level fields into new fields.

You can use dot notation (embedded.field.name) to clone embedded frame fields.

Only applicable to datasets that contain videos.

Parameters: field_mapping – a dict mapping field names to new field names into which to clone each field

clear_sample_field(field_name)¶

Clears the values of the field from all samples in the dataset.

The field will remain in the dataset’s schema, and all samples will have the value None for the field.

You can use dot notation (embedded.field.name) to clear embedded fields.

Parameters: field_name – the field name or embedded.field.name

clear_sample_fields(field_names)¶

Clears the values of the fields from all samples in the dataset.

The field will remain in the dataset’s schema, and all samples will have the value None for the field.

You can use dot notation (embedded.field.name) to clear embedded fields.

Parameters: field_names – the field name or iterable of field names

clear_frame_field(field_name)¶

Clears the values of the frame-level field from all samples in the dataset.

The field will remain in the dataset’s frame schema, and all frames will have the value None for the field.

You can use dot notation (embedded.field.name) to clear embedded frame fields.

Only applicable to datasets that contain videos.

Parameters: field_name – the field name or embedded.field.name

clear_frame_fields(field_names)¶

Clears the values of the frame-level fields from all samples in the dataset.

The fields will remain in the dataset’s frame schema, and all frames will have the value None for the field.

You can use dot notation (embedded.field.name) to clear embedded frame fields.

Only applicable to datasets that contain videos.

Parameters: field_names – the field name or iterable of field names

delete_sample_field(field_name, error_level=0)¶

Deletes the field from all samples in the dataset.

You can use dot notation (embedded.field.name) to delete embedded fields.

Parameters

field_name – the field name or embedded.field.name
error_level (0) – the error level to use. Valid values are:
0 (-) – raise error if a top-level field cannot be deleted
1 (-) – log warning if a top-level field cannot be deleted
2 (-) – ignore top-level fields that cannot be deleted

delete_sample_fields(field_names, error_level=0)¶

Deletes the fields from all samples in the dataset.

You can use dot notation (embedded.field.name) to delete embedded fields.

Parameters

field_names – the field name or iterable of field names
error_level (0) – the error level to use. Valid values are:
0 (-) – raise error if a top-level field cannot be deleted
1 (-) – log warning if a top-level field cannot be deleted
2 (-) – ignore top-level fields that cannot be deleted

remove_dynamic_sample_field(field_name, error_level=0)¶

Removes the dynamic embedded sample field from the dataset’s schema.

The underlying data is not deleted from the samples.

Parameters

field_name – the embedded.field.name
error_level (0) – the error level to use. Valid values are:
0 (-) – raise error if a top-level field cannot be removed
1 (-) – log warning if a top-level field cannot be removed
2 (-) – ignore top-level fields that cannot be removed

remove_dynamic_sample_fields(field_names, error_level=0)¶

Removes the dynamic embedded sample fields from the dataset’s schema.

The underlying data is not deleted from the samples.

Parameters

field_names – the embedded.field.name or iterable of field names
error_level (0) – the error level to use. Valid values are:
0 (-) – raise error if a top-level field cannot be removed
1 (-) – log warning if a top-level field cannot be removed
2 (-) – ignore top-level fields that cannot be removed

delete_frame_field(field_name, error_level=0)¶

Deletes the frame-level field from all samples in the dataset.

You can use dot notation (embedded.field.name) to delete embedded frame fields.

Only applicable to datasets that contain videos.

Parameters

field_name – the field name or embedded.field.name
error_level (0) – the error level to use. Valid values are:
0 (-) – raise error if a top-level field cannot be deleted
1 (-) – log warning if a top-level field cannot be deleted
2 (-) – ignore top-level fields that cannot be deleted

delete_frame_fields(field_names, error_level=0)¶

Deletes the frame-level fields from all samples in the dataset.

You can use dot notation (embedded.field.name) to delete embedded frame fields.

Only applicable to datasets that contain videos.

Parameters

field_names – a field name or iterable of field names
error_level (0) – the error level to use. Valid values are:
0 (-) – raise error if a top-level field cannot be deleted
1 (-) – log warning if a top-level field cannot be deleted
2 (-) – ignore top-level fields that cannot be deleted

remove_dynamic_frame_field(field_name, error_level=0)¶

Removes the dynamic embedded frame field from the dataset’s schema.

The underlying data is not deleted from the frames.

Parameters

field_name – the embedded.field.name
error_level (0) – the error level to use. Valid values are:
0 (-) – raise error if a top-level field cannot be removed
1 (-) – log warning if a top-level field cannot be removed
2 (-) – ignore top-level fields that cannot be removed

remove_dynamic_frame_fields(field_names, error_level=0)¶

Removes the dynamic embedded frame fields from the dataset’s schema.

The underlying data is not deleted from the frames.

Parameters

field_names – the embedded.field.name or iterable of field names
error_level (0) – the error level to use. Valid values are:
0 (-) – raise error if a top-level field cannot be removed
1 (-) – log warning if a top-level field cannot be removed
2 (-) – ignore top-level fields that cannot be removed

add_group_slice(name, media_type)¶

Adds a group slice with the given media type to the dataset, if necessary.

Parameters

name – a group slice name
media_type – the media type of the slice

rename_group_slice(name, new_name)¶

Renames the group slice with the given name.

Parameters

name – the group slice name
new_name – the new group slice name

delete_group_slice(name)¶

Deletes all samples in the given group slice from the dataset.

Parameters: name – a group slice name

iter_samples(progress=False, autosave=False, batch_size=None, batching_strategy=None)¶

Returns an iterator over the samples in the dataset.

Examples:

import random as r
import string as s

import fiftyone as fo
import fiftyone.zoo as foz

dataset = foz.load_zoo_dataset("cifar10", split="test")

def make_label():
    return "".join(r.choice(s.ascii_letters) for i in range(10))

# No save context
for sample in dataset.iter_samples(progress=True):
    sample.ground_truth.label = make_label()
    sample.save()

# Save using default batching strategy
for sample in dataset.iter_samples(progress=True, autosave=True):
    sample.ground_truth.label = make_label()

# Save in batches of 10
for sample in dataset.iter_samples(
    progress=True, autosave=True, batch_size=10
):
    sample.ground_truth.label = make_label()

# Save every 0.5 seconds
for sample in dataset.iter_samples(
    progress=True, autosave=True, batch_size=0.5
):
    sample.ground_truth.label = make_label()

Parameters

progress (False) – whether to render a progress bar (True/False), use the default value fiftyone.config.show_progress_bars (None), or a progress callback function to invoke instead
autosave (False) – whether to automatically save changes to samples emitted by this iterator
batch_size (None) – the batch size to use when autosaving samples. If a batching_strategy is provided, this parameter configures the strategy as described below. If no batching_strategy is provided, this can either be an integer specifying the number of samples to save in a batch (in which case batching_strategy is implicitly set to "static") or a float number of seconds between batched saves (in which case batching_strategy is implicitly set to "latency")
batching_strategy (None) –
the batching strategy to use for each save operation when autosaving samples. Supported values are:
- "static": a fixed sample batch size for each save
- "size": a target batch size, in bytes, for each save
- "latency": a target latency, in seconds, between saves
By default, fo.config.default_batcher is used

Returns

an iterator over fiftyone.core.sample.Sample instances

iter_groups(group_slices=None, progress=False, autosave=False, batch_size=None, batching_strategy=None)¶

Returns an iterator over the groups in the dataset.

Examples:

import random as r
import string as s

import fiftyone as fo
import fiftyone.zoo as foz

dataset = foz.load_zoo_dataset("quickstart-groups")

def make_label():
    return "".join(r.choice(s.ascii_letters) for i in range(10))

# No save context
for group in dataset.iter_groups(progress=True):
    for sample in group.values():
        sample["test"] = make_label()
        sample.save()

# Save using default batching strategy
for group in dataset.iter_groups(progress=True, autosave=True):
    for sample in group.values():
        sample["test"] = make_label()

# Save in batches of 10
for group in dataset.iter_groups(
    progress=True, autosave=True, batch_size=10
):
    for sample in group.values():
        sample["test"] = make_label()

# Save every 0.5 seconds
for group in dataset.iter_groups(
    progress=True, autosave=True, batch_size=0.5
):
    for sample in group.values():
        sample["test"] = make_label()

Parameters

group_slices (None) – an optional subset of group slices to load
progress (False) – whether to render a progress bar (True/False), use the default value fiftyone.config.show_progress_bars (None), or a progress callback function to invoke instead
autosave (False) – whether to automatically save changes to samples emitted by this iterator
batch_size (None) – the batch size to use when autosaving samples. If a batching_strategy is provided, this parameter configures the strategy as described below. If no batching_strategy is provided, this can either be an integer specifying the number of samples to save in a batch (in which case batching_strategy is implicitly set to "static") or a float number of seconds between batched saves (in which case batching_strategy is implicitly set to "latency")
batching_strategy (None) –
the batching strategy to use for each save operation when autosaving samples. Supported values are:
- "static": a fixed sample batch size for each save
- "size": a target batch size, in bytes, for each save
- "latency": a target latency, in seconds, between saves
By default, fo.config.default_batcher is used

Returns

an iterator that emits dicts mapping group slice names to fiftyone.core.sample.Sample instances, one per group

get_group(group_id, group_slices=None)¶

Returns a dict containing the samples for the given group ID.

Examples:

import fiftyone as fo
import fiftyone.zoo as foz

dataset = foz.load_zoo_dataset("quickstart-groups")

group_id = dataset.take(1).first().group.id
group = dataset.get_group(group_id)

print(group.keys())
# ['left', 'right', 'pcd']

Parameters

group_id – a group ID
group_slices (None) – an optional subset of group slices to load

Returns

a dict mapping group names to fiftyone.core.sample.Sample instances

Raises

KeyError – if the group ID is not found

add_sample(sample, expand_schema=True, dynamic=False, validate=True)¶

Adds the given sample to the dataset.

If the sample instance does not belong to a dataset, it is updated in-place to reflect its membership in this dataset. If the sample instance belongs to another dataset, it is not modified.

Parameters

sample – a fiftyone.core.sample.Sample
expand_schema (True) – whether to dynamically add new sample fields encountered to the dataset schema. If False, an error is raised if the sample’s schema is not a subset of the dataset schema
dynamic (False) – whether to declare dynamic attributes of embedded document fields that are encountered
validate (True) – whether to validate that the fields of the sample are compliant with the dataset schema before adding it

Returns

the ID of the sample in the dataset

add_samples(samples, expand_schema=True, dynamic=False, validate=True, progress=None, num_samples=None)¶

Adds the given samples to the dataset.

Any sample instances that do not belong to a dataset are updated in-place to reflect membership in this dataset. Any sample instances that belong to other datasets are not modified.

Parameters

samples – an iterable of fiftyone.core.sample.Sample instances or a fiftyone.core.collections.SampleCollection
expand_schema (True) – whether to dynamically add new sample fields encountered to the dataset schema. If False, an error is raised if a sample’s schema is not a subset of the dataset schema
dynamic (False) – whether to declare dynamic attributes of embedded document fields that are encountered
validate (True) – whether to validate that the fields of each sample are compliant with the dataset schema before adding it
progress (None) – whether to render a progress bar (True/False), use the default value fiftyone.config.show_progress_bars (None), or a progress callback function to invoke instead
num_samples (None) – the number of samples in samples. If not provided, this is computed (if possible) via len(samples) if needed for progress tracking

Returns

a list of IDs of the samples in the dataset

add_collection(sample_collection, include_info=True, overwrite_info=False, new_ids=False, progress=None)¶

Adds the contents of the given collection to the dataset.

This method is a special case of Dataset.merge_samples() that adds samples with new IDs to this dataset and omits any samples with existing IDs (the latter would only happen in rare cases).

Use Dataset.merge_samples() if you have multiple datasets whose samples refer to the same source media.

Parameters

sample_collection – a fiftyone.core.collections.SampleCollection
include_info (True) – whether to merge dataset-level information such as info and classes
overwrite_info (False) – whether to overwrite existing dataset-level information. Only applicable when include_info is True
new_ids (False) – whether to generate new sample/frame/group IDs. By default, the IDs of the input collection are retained
progress (None) – whether to render a progress bar (True/False), use the default value fiftyone.config.show_progress_bars (None), or a progress callback function to invoke instead

Returns

a list of IDs of the samples that were added to this dataset

merge_sample(sample, key_field='filepath', skip_existing=False, insert_new=True, fields=None, omit_fields=None, merge_lists=True, overwrite=True, expand_schema=True, validate=True, dynamic=False)¶

Merges the fields of the given sample into this dataset.

By default, the sample is merged with an existing sample with the same absolute filepath, if one exists. Otherwise a new sample is inserted. You can customize this behavior via the key_field, skip_existing, and insert_new parameters.

The behavior of this method is highly customizable. By default, all top-level fields from the provided sample are merged in, overwriting any existing values for those fields, with the exception of list fields (e.g., tags) and label list fields (e.g., fiftyone.core.labels.Detections fields), in which case the elements of the lists themselves are merged. In the case of label list fields, labels with the same id in both samples are updated rather than duplicated.

To avoid confusion between missing fields and fields whose value is None, None-valued fields are always treated as missing while merging.

This method can be configured in numerous ways, including:

Whether new fields can be added to the dataset schema
Whether list fields should be treated as ordinary fields and merged as a whole rather than merging their elements
Whether to merge only specific fields, or all but certain fields
Mapping input sample fields to different field names of this sample

Parameters

sample – a fiftyone.core.sample.Sample
key_field ("filepath") – the sample field to use to decide whether to join with an existing sample
skip_existing (False) – whether to skip existing samples (True) or merge them (False)
insert_new (True) – whether to insert new samples (True) or skip them (False)
fields (None) – an optional field or iterable of fields to which to restrict the merge. May contain frame fields for video samples. This can also be a dict mapping field names of the input sample to field names of this dataset
omit_fields (None) – an optional field or iterable of fields to exclude from the merge. May contain frame fields for video samples
merge_lists (True) – whether to merge the elements of list fields (e.g., tags) and label list fields (e.g., fiftyone.core.labels.Detections fields) rather than merging the entire top-level field like other field types. For label lists fields, existing fiftyone.core.label.Label elements are either replaced (when overwrite is True) or kept (when overwrite is False) when their id matches a label from the provided sample
overwrite (True) – whether to overwrite (True) or skip (False) existing fields and label elements
expand_schema (True) – whether to dynamically add new fields encountered to the dataset schema. If False, an error is raised if any fields are not in the dataset schema
validate (True) – whether to validate values for existing fields
dynamic (False) – whether to declare dynamic embedded document fields

merge_samples(samples, key_field='filepath', key_fcn=None, skip_existing=False, insert_new=True, fields=None, omit_fields=None, merge_lists=True, overwrite=True, expand_schema=True, dynamic=False, include_info=True, overwrite_info=False, progress=None, num_samples=None)¶

Merges the given samples into this dataset.

Note

This method requires the ability to create unique indexes on the key_field of each collection.

See add_collection() if you want to add samples from one collection to another dataset without a uniqueness constraint.

By default, samples with the same absolute filepath are merged, but you can customize this behavior via the key_field and key_fcn parameters. For example, you could set key_fcn = lambda sample: os.path.basename(sample.filepath) to merge samples with the same base filename.

The behavior of this method is highly customizable. By default, all top-level fields from the provided samples are merged in, overwriting any existing values for those fields, with the exception of list fields (e.g., tags) and label list fields (e.g., fiftyone.core.labels.Detections fields), in which case the elements of the lists themselves are merged. In the case of label list fields, labels with the same id in both collections are updated rather than duplicated.

To avoid confusion between missing fields and fields whose value is None, None-valued fields are always treated as missing while merging.

This method can be configured in numerous ways, including:

Whether existing samples should be modified or skipped
Whether new samples should be added or omitted
Whether new fields can be added to the dataset schema
Whether list fields should be treated as ordinary fields and merged as a whole rather than merging their elements
Whether to merge only specific fields, or all but certain fields
Mapping input fields to different field names of this dataset

Parameters

samples – a fiftyone.core.collections.SampleCollection or iterable of fiftyone.core.sample.Sample instances
key_field ("filepath") – the sample field to use to decide whether to join with an existing sample
key_fcn (None) – a function that accepts a fiftyone.core.sample.Sample instance and computes a key to decide if two samples should be merged. If a key_fcn is provided, key_field is ignored
skip_existing (False) – whether to skip existing samples (True) or merge them (False)
insert_new (True) – whether to insert new samples (True) or skip them (False)
fields (None) – an optional field or iterable of fields to which to restrict the merge. If provided, fields other than these are omitted from samples when merging or adding samples. One exception is that filepath is always included when adding new samples, since the field is required. This can also be a dict mapping field names of the input collection to field names of this dataset
omit_fields (None) – an optional field or iterable of fields to exclude from the merge. If provided, these fields are omitted from samples, if present, when merging or adding samples. One exception is that filepath is always included when adding new samples, since the field is required
merge_lists (True) – whether to merge the elements of list fields (e.g., tags) and label list fields (e.g., fiftyone.core.labels.Detections fields) rather than merging the entire top-level field like other field types. For label lists fields, existing fiftyone.core.label.Label elements are either replaced (when overwrite is True) or kept (when overwrite is False) when their id matches a label from the provided samples
overwrite (True) – whether to overwrite (True) or skip (False) existing fields and label elements
expand_schema (True) – whether to dynamically add new fields encountered to the dataset schema. If False, an error is raised if a sample’s schema is not a subset of the dataset schema
dynamic (False) – whether to declare dynamic attributes of embedded document fields that are encountered. Only applicable when samples is not a fiftyone.core.collections.SampleCollection
include_info (True) – whether to merge dataset-level information such as info and classes. Only applicable when samples is a fiftyone.core.collections.SampleCollection
overwrite_info (False) – whether to overwrite existing dataset-level information. Only applicable when samples is a fiftyone.core.collections.SampleCollection and include_info is True
progress (None) – whether to render a progress bar (True/False), use the default value fiftyone.config.show_progress_bars (None), or a progress callback function to invoke instead
num_samples (None) – the number of samples in samples. If not provided, this is computed (if possible) via len(samples) if needed for progress tracking

delete_samples(samples_or_ids)¶

Deletes the given sample(s) from the dataset.

If reference to a sample exists in memory, the sample will be updated such that sample.in_dataset is False.

Parameters

samples_or_ids –

the sample(s) to delete. Can be any of the following:

a sample ID
an iterable of sample IDs
a fiftyone.core.sample.Sample or fiftyone.core.sample.SampleView
an iterable of fiftyone.core.sample.Sample or fiftyone.core.sample.SampleView instances
a fiftyone.core.collections.SampleCollection

delete_frames(frames_or_ids)¶

Deletes the given frames(s) from the dataset.

If reference to a frame exists in memory, the frame will be updated such that frame.in_dataset is False.

Parameters

frames_or_ids –

the frame(s) to delete. Can be any of the following:

a frame ID
an iterable of frame IDs
a fiftyone.core.frame.Frame or fiftyone.core.frame.FrameView
a fiftyone.core.sample.Sample or fiftyone.core.sample.SampleView whose frames to delete
an iterable of fiftyone.core.frame.Frame or fiftyone.core.frame.FrameView instances
an iterable of fiftyone.core.sample.Sample or fiftyone.core.sample.SampleView instances whose frames to delete
a fiftyone.core.collections.SampleCollection whose frames to delete

delete_groups(groups_or_ids)¶

Deletes the given groups(s) from the dataset.

If reference to a sample exists in memory, the sample will be updated such that sample.in_dataset is False.

Parameters

groups_or_ids –

the group(s) to delete. Can be any of the following:

a group ID
an iterable of group IDs
a fiftyone.core.sample.Sample or fiftyone.core.sample.SampleView
a group dict returned by get_group()
an iterable of fiftyone.core.sample.Sample or fiftyone.core.sample.SampleView instances
an iterable of group dicts returned by get_group()
a fiftyone.core.collections.SampleCollection

delete_labels(labels=None, ids=None, tags=None, view=None, fields=None)¶

Deletes the specified labels from the dataset.

You can specify the labels to delete via any of the following methods:

Provide the labels argument, which should contain a list of dicts in the format returned by fiftyone.core.session.Session.selected_labels
Provide the ids or tags arguments to specify the labels to delete via their IDs and/or tags
Provide the view argument to delete all of the labels in a view into this dataset. This syntax is useful if you have constructed a fiftyone.core.view.DatasetView defining the labels to delete

Additionally, you can specify the fields argument to restrict deletion to specific field(s), either for efficiency or to ensure that labels from other fields are not deleted if their contents are included in the other arguments.

Parameters

labels (None) – a list of dicts specifying the labels to delete in the format returned by fiftyone.core.session.Session.selected_labels
ids (None) – an ID or iterable of IDs of the labels to delete
tags (None) – a tag or iterable of tags of the labels to delete
view (None) – a fiftyone.core.view.DatasetView into this dataset containing the labels to delete
fields (None) – a field or iterable of fields from which to delete labels

save()¶

Saves the dataset to the database.

This only needs to be called when dataset-level information such as its Dataset.info() is modified.

property has_saved_views¶: Whether this dataset has any saved views.

has_saved_view(name)¶

Whether this dataset has a saved view with the given name.

Parameters: name – a saved view name
Returns: True/False

list_saved_views(info=False)¶

List saved views on this dataset.

Parameters: info (False) – whether to return info dicts describing each saved view rather than just their names
Returns: a list of saved view names or info dicts

save_view(name, view, description=None, color=None, overwrite=False)¶

Saves the given view into this dataset under the given name so it can be loaded later via load_saved_view().

Examples:

import fiftyone as fo
import fiftyone.zoo as foz
from fiftyone import ViewField as F

dataset = foz.load_zoo_dataset("quickstart")
view = dataset.filter_labels("ground_truth", F("label") == "cat")

dataset.save_view("cats", view)

also_view = dataset.load_saved_view("cats")
assert view == also_view

Parameters

name – a name for the saved view
view – a fiftyone.core.view.DatasetView
description (None) – an optional string description
color (None) – an optional RGB hex string like '#FF6D04'
overwrite (False) – whether to overwrite an existing saved view with the same name

get_saved_view_info(name)¶

Loads the editable information about the saved view with the given name.

Examples:

import fiftyone as fo
import fiftyone.zoo as foz

dataset = foz.load_zoo_dataset("quickstart")

view = dataset.limit(10)
dataset.save_view("test", view)

print(dataset.get_saved_view_info("test"))

Parameters: name – the name of a saved view
Returns: a dict of editable info

update_saved_view_info(name, info)¶

Updates the editable information for the saved view with the given name.

Examples:

import fiftyone as fo
import fiftyone.zoo as foz

dataset = foz.load_zoo_dataset("quickstart")

view = dataset.limit(10)
dataset.save_view("test", view)

# Update the saved view's name and add a description
info = dict(
    name="a new name",
    description="a description",
)
dataset.update_saved_view_info("test", info)

Parameters

name – the name of a saved view
info – a dict whose keys are a subset of the keys returned by get_saved_view_info()

load_saved_view(name)¶

Loads the saved view with the given name.

Examples:

import fiftyone as fo
import fiftyone.zoo as foz
from fiftyone import ViewField as F

dataset = foz.load_zoo_dataset("quickstart")
view = dataset.filter_labels("ground_truth", F("label") == "cat")

dataset.save_view("cats", view)

also_view = dataset.load_saved_view("cats")
assert view == also_view

Parameters: name – the name of a saved view
Returns: a fiftyone.core.view.DatasetView

delete_saved_view(name)¶

Deletes the saved view with the given name.

Parameters: name – the name of a saved view

delete_saved_views()¶: Deletes all saved views from this dataset.

property has_workspaces¶: Whether this dataset has any saved workspaces.

has_workspace(name)¶

Whether this dataset has a saved workspace with the given name.

Parameters: name – a saved workspace name
Returns: True/False

list_workspaces(info=False)¶

List saved workspaces on this dataset.

Parameters: info (False) – whether to return info dicts describing each saved workspace rather than just their names
Returns: a list of saved workspace names or info dicts

save_workspace(name, workspace, description=None, color=None, overwrite=False)¶

Saves a workspace into this dataset under the given name so it can be loaded later via load_workspace().

Examples:

import fiftyone as fo
import fiftyone.zoo as foz

dataset = foz.load_zoo_dataset("quickstart")

embeddings_panel = fo.Panel(
    type="Embeddings",
    state=dict(
        brainResult="img_viz",
        colorByField="metadata.size_bytes"
    ),
)
workspace = fo.Space(children=[embeddings_panel])

workspace_name = "embeddings-workspace"
description = "Show embeddings only"
dataset.save_workspace(
    workspace_name,
    workspace,
    description=description
)
assert dataset.has_workspace(workspace_name)

also_workspace = dataset.load_workspace(workspace_name)
assert workspace == also_workspace

Parameters

name – a name for the saved workspace
workspace – a fiftyone.core.odm.workspace.Space
description (None) – an optional string description
color (None) – an optional RGB hex string like '#FF6D04'
overwrite (False) – whether to overwrite an existing workspace with the same name

Raises

ValueError – if overwrite==False and workspace with name already exists

load_workspace(name)¶

Loads the saved workspace with the given name.

Examples:

import fiftyone as fo
import fiftyone.zoo as foz

dataset = foz.load_zoo_dataset("quickstart")

embeddings_panel = fo.Panel(
    type="Embeddings",
    state=dict(brainResult="img_viz", colorByField="metadata.size_bytes"),
)
workspace = fo.Space(children=[embeddings_panel])
workspace_name = "embeddings-workspace"
dataset.save_workspace(workspace_name, workspace)

# Some time later ... load the workspace
loaded_workspace = dataset.load_workspace(workspace_name)
assert workspace == loaded_workspace

# Launch app with the loaded workspace!
session = fo.launch_app(dataset, spaces=loaded_workspace)

# Or set via session later on
session.spaces = loaded_workspace

Parameters: name – the name of a saved workspace
Returns: a fiftyone.core.odm.workspace.Space
Raises: ValueError – if name is not a saved workspace

get_workspace_info(name)¶

Gets the information about the workspace with the given name.

Examples:

import fiftyone as fo
import fiftyone.zoo as foz

dataset = foz.load_zoo_dataset("quickstart")

workspace = fo.Space()
description = "A really cool (apparently empty?) workspace"
dataset.save_workspace("test", workspace, description=description)

print(dataset.get_workspace_info("test"))

Parameters: name – the name of a saved view
Returns: a dict of editable info

update_workspace_info(name, info)¶

Updates the editable information for the saved view with the given name.

Examples:

import fiftyone as fo
import fiftyone.zoo as foz

dataset = foz.load_zoo_dataset("quickstart")

workspace = fo.Space()
dataset.save_workspace("test", view)

# Update the workspace's name and add a description, color
info = dict(
    name="a new name",
    color="#FF6D04",
    description="a description",
)
dataset.update_workspace_info("test", info)

Parameters

name – the name of a saved workspace
info – a dict whose keys are a subset of the keys returned by get_workspace_info()

delete_workspace(name)¶

Deletes the saved workspace with the given name.

Parameters: name – the name of a saved workspace
Raises: ValueError – if name is not a saved workspace

delete_workspaces()¶: Deletes all saved workspaces from this dataset.

clone(name=None, persistent=False)¶

Creates a copy of the dataset.

Dataset clones contain deep copies of all samples and dataset-level information in the source dataset. The source media files, however, are not copied.

Parameters

name (None) – a name for the cloned dataset. By default, get_default_dataset_name() is used
persistent (False) – whether the cloned dataset should be persistent

Returns

the new Dataset

clear()¶

Removes all samples from the dataset.

If reference to a sample exists in memory, the sample will be updated such that sample.in_dataset is False.

clear_frames()¶

Removes all frame labels from the dataset.

If reference to a frame exists in memory, the frame will be updated such that frame.in_dataset is False.

ensure_frames()¶

Ensures that the video dataset contains frame instances for every frame of each sample’s source video.

Empty frames will be inserted for missing frames, and already existing frames are left unchanged.

delete()¶

Deletes the dataset.

Once deleted, only the name and deleted attributes of a dataset may be accessed.

If reference to a sample exists in memory, the sample will be updated such that sample.in_dataset is False.

add_dir(dataset_dir=None, dataset_type=None, data_path=None, labels_path=None, label_field=None, tags=None, expand_schema=True, dynamic=False, add_info=True, progress=None, **kwargs)¶

Adds the contents of the given directory to the dataset.

You can perform imports with this method via the following basic patterns:

Provide dataset_dir and dataset_type to import the contents of a directory that is organized in the default layout for the dataset type as documented in this guide
Provide dataset_type along with data_path, labels_path, or other type-specific parameters to perform a customized import. This syntax provides the flexibility to, for example, perform labels-only imports or imports where the source media lies in a different location than the labels

In either workflow, the remaining parameters of this method can be provided to further configure the import.

See this guide for example usages of this method and descriptions of the available dataset types.

Parameters

dataset_dir (None) – the dataset directory. This can be omitted for certain dataset formats if you provide arguments such as data_path and labels_path
dataset_type (None) – the fiftyone.types.Dataset type of the dataset
data_path (None) –
an optional parameter that enables explicit control over the location of the media for certain dataset types. Can be any of the following:
- a folder name like "data" or "data/" specifying a subfolder of dataset_dir in which the media lies
- an absolute directory path in which the media lies. In this case, the dataset_dir has no effect on the location of the data
- a filename like "data.json" specifying the filename of a JSON manifest file in dataset_dir that maps UUIDs to media filepaths. Files of this format are generated when passing the export_media="manifest" option to fiftyone.core.collections.SampleCollection.export()
- an absolute filepath to a JSON manifest file. In this case, dataset_dir has no effect on the location of the data
- a dict mapping filenames to absolute filepaths
By default, it is assumed that the data can be located in the default location within dataset_dir for the dataset type
labels_path (None) –
an optional parameter that enables explicit control over the location of the labels. Only applicable when importing certain labeled dataset formats. Can be any of the following:
- a type-specific folder name like "labels" or "labels/" or a filename like "labels.json" or "labels.xml" specifying the location in dataset_dir of the labels file(s)
- an absolute directory or filepath containing the labels file(s). In this case, dataset_dir has no effect on the location of the labels
For labeled datasets, this parameter defaults to the location in dataset_dir of the labels for the default layout of the dataset type being imported
label_field (None) – controls the field(s) in which imported labels are stored. Only applicable if dataset_importer is a fiftyone.utils.data.importers.LabeledImageDatasetImporter or fiftyone.utils.data.importers.LabeledVideoDatasetImporter. If the importer produces a single fiftyone.core.labels.Label instance per sample/frame, this argument specifies the name of the field to use; the default is "ground_truth". If the importer produces a dictionary of labels per sample, this argument can be either a string prefix to prepend to each label key or a dict mapping label keys to field names; the default in this case is to directly use the keys of the imported label dictionaries as field names
tags (None) – an optional tag or iterable of tags to attach to each sample
expand_schema (True) – whether to dynamically add new sample fields encountered to the dataset schema. If False, an error is raised if a sample’s schema is not a subset of the dataset schema
dynamic (False) – whether to declare dynamic attributes of embedded document fields that are encountered
add_info (True) – whether to add dataset info from the importer (if any) to the dataset’s info
progress (None) – whether to render a progress bar (True/False), use the default value fiftyone.config.show_progress_bars (None), or a progress callback function to invoke instead
**kwargs – optional keyword arguments to pass to the constructor of the fiftyone.utils.data.importers.DatasetImporter for the specified dataset_type

Returns

a list of IDs of the samples that were added to the dataset

merge_dir(dataset_dir=None, dataset_type=None, data_path=None, labels_path=None, label_field=None, tags=None, key_field='filepath', key_fcn=None, skip_existing=False, insert_new=True, fields=None, omit_fields=None, merge_lists=True, overwrite=True, expand_schema=True, dynamic=False, add_info=True, progress=None, **kwargs)¶

Merges the contents of the given directory into the dataset.

Note

This method requires the ability to create unique indexes on the key_field of each collection.

See add_dir() if you want to add samples without a uniqueness constraint.

You can perform imports with this method via the following basic patterns:

Provide dataset_dir and dataset_type to import the contents of a directory that is organized in the default layout for the dataset type as documented in this guide
Provide dataset_type along with data_path, labels_path, or other type-specific parameters to perform a customized import. This syntax provides the flexibility to, for example, perform labels-only imports or imports where the source media lies in a different location than the labels

In either workflow, the remaining parameters of this method can be provided to further configure the import.

See this guide for example usages of this method and descriptions of the available dataset types.

By default, samples with the same absolute filepath are merged, but you can customize this behavior via the key_field and key_fcn parameters. For example, you could set key_fcn = lambda sample: os.path.basename(sample.filepath) to merge samples with the same base filename.

The behavior of this method is highly customizable. By default, all top-level fields from the imported samples are merged in, overwriting any existing values for those fields, with the exception of list fields (e.g., tags) and label list fields (e.g., fiftyone.core.labels.Detections fields), in which case the elements of the lists themselves are merged. In the case of label list fields, labels with the same id in both collections are updated rather than duplicated.

To avoid confusion between missing fields and fields whose value is None, None-valued fields are always treated as missing while merging.

This method can be configured in numerous ways, including:

Whether existing samples should be modified or skipped
Whether new samples should be added or omitted
Whether new fields can be added to the dataset schema
Whether list fields should be treated as ordinary fields and merged as a whole rather than merging their elements
Whether to merge only specific fields, or all but certain fields
Mapping input fields to different field names of this dataset

Parameters

dataset_dir (None) – the dataset directory. This can be omitted for certain dataset formats if you provide arguments such as data_path and labels_path
dataset_type (None) – the fiftyone.types.Dataset type of the dataset
data_path (None) –
an optional parameter that enables explicit control over the location of the media for certain dataset types. Can be any of the following:
- a folder name like "data" or "data/" specifying a subfolder of dataset_dir in which the media lies
- an absolute directory path in which the media lies. In this case, the dataset_dir has no effect on the location of the data
- a filename like "data.json" specifying the filename of a JSON manifest file in dataset_dir that maps UUIDs to media filepaths. Files of this format are generated when passing the export_media="manifest" option to fiftyone.core.collections.SampleCollection.export()
- an absolute filepath to a JSON manifest file. In this case, dataset_dir has no effect on the location of the data
- a dict mapping filenames to absolute filepaths
By default, it is assumed that the data can be located in the default location within dataset_dir for the dataset type
labels_path (None) –
an optional parameter that enables explicit control over the location of the labels. Only applicable when importing certain labeled dataset formats. Can be any of the following:
- a type-specific folder name like "labels" or "labels/" or a filename like "labels.json" or "labels.xml" specifying the location in dataset_dir of the labels file(s)
- an absolute directory or filepath containing the labels file(s). In this case, dataset_dir has no effect on the location of the labels
For labeled datasets, this parameter defaults to the location in dataset_dir of the labels for the default layout of the dataset type being imported
label_field (None) – controls the field(s) in which imported labels are stored. Only applicable if dataset_importer is a fiftyone.utils.data.importers.LabeledImageDatasetImporter or fiftyone.utils.data.importers.LabeledVideoDatasetImporter. If the importer produces a single fiftyone.core.labels.Label instance per sample/frame, this argument specifies the name of the field to use; the default is "ground_truth". If the importer produces a dictionary of labels per sample, this argument can be either a string prefix to prepend to each label key or a dict mapping label keys to field names; the default in this case is to directly use the keys of the imported label dictionaries as field names
tags (None) – an optional tag or iterable of tags to attach to each sample
key_field ("filepath") – the sample field to use to decide whether to join with an existing sample
key_fcn (None) – a function that accepts a fiftyone.core.sample.Sample instance and computes a key to decide if two samples should be merged. If a key_fcn is provided, key_field is ignored
skip_existing (False) – whether to skip existing samples (True) or merge them (False)
insert_new (True) – whether to insert new samples (True) or skip them (False)
fields (None) – an optional field or iterable of fields to which to restrict the merge. If provided, fields other than these are omitted from samples when merging or adding samples. One exception is that filepath is always included when adding new samples, since the field is required. This can also be a dict mapping field names of the input collection to field names of this dataset
omit_fields (None) – an optional field or iterable of fields to exclude from the merge. If provided, these fields are omitted from imported samples, if present. One exception is that filepath is always included when adding new samples, since the field is required
merge_lists (True) – whether to merge the elements of list fields (e.g., tags) and label list fields (e.g., fiftyone.core.labels.Detections fields) rather than merging the entire top-level field like other field types. For label lists fields, existing fiftyone.core.label.Label elements are either replaced (when overwrite is True) or kept (when overwrite is False) when their id matches a label from the provided samples
overwrite (True) – whether to overwrite (True) or skip (False) existing fields and label elements
expand_schema (True) – whether to dynamically add new fields encountered to the dataset schema. If False, an error is raised if a sample’s schema is not a subset of the dataset schema
dynamic (False) – whether to declare dynamic attributes of embedded document fields that are encountered
add_info (True) – whether to add dataset info from the importer (if any) to the dataset
progress (None) – whether to render a progress bar (True/False), use the default value fiftyone.config.show_progress_bars (None), or a progress callback function to invoke instead
**kwargs – optional keyword arguments to pass to the constructor of the fiftyone.utils.data.importers.DatasetImporter for the specified dataset_type

add_archive(archive_path, dataset_type=None, data_path=None, labels_path=None, label_field=None, tags=None, expand_schema=True, dynamic=False, add_info=True, cleanup=True, progress=None, **kwargs)¶

Adds the contents of the given archive to the dataset.

If a directory with the same root name as archive_path exists, it is assumed that this directory contains the extracted contents of the archive, and thus the archive is not re-extracted.

See this guide for example usages of this method and descriptions of the available dataset types.

Note

The following archive formats are explicitly supported:

.zip, .tar, .tar.gz, .tgz, .tar.bz, .tbz

If an archive not in the above list is found, extraction will be attempted via the patool package, which supports many formats but may require that additional system packages be installed.

Parameters

archive_path – the path to an archive of a dataset directory
dataset_type (None) – the fiftyone.types.Dataset type of the dataset in archive_path
data_path (None) –
an optional parameter that enables explicit control over the location of the media for certain dataset types. Can be any of the following:
- a folder name like "data" or "data/" specifying a subfolder of dataset_dir in which the media lies
- an absolute directory path in which the media lies. In this case, the archive_path has no effect on the location of the data
- a filename like "data.json" specifying the filename of a JSON manifest file in archive_path that maps UUIDs to media filepaths. Files of this format are generated when passing the export_media="manifest" option to fiftyone.core.collections.SampleCollection.export()
- an absolute filepath to a JSON manifest file. In this case, archive_path has no effect on the location of the data
- a dict mapping filenames to absolute filepaths
By default, it is assumed that the data can be located in the default location within archive_path for the dataset type
labels_path (None) –
an optional parameter that enables explicit control over the location of the labels. Only applicable when importing certain labeled dataset formats. Can be any of the following:
- a type-specific folder name like "labels" or "labels/" or a filename like "labels.json" or "labels.xml" specifying the location in archive_path of the labels file(s)
- an absolute directory or filepath containing the labels file(s). In this case, archive_path has no effect on the location of the labels
For labeled datasets, this parameter defaults to the location in archive_path of the labels for the default layout of the dataset type being imported
label_field (None) – controls the field(s) in which imported labels are stored. Only applicable if dataset_importer is a fiftyone.utils.data.importers.LabeledImageDatasetImporter or fiftyone.utils.data.importers.LabeledVideoDatasetImporter. If the importer produces a single fiftyone.core.labels.Label instance per sample/frame, this argument specifies the name of the field to use; the default is "ground_truth". If the importer produces a dictionary of labels per sample, this argument can be either a string prefix to prepend to each label key or a dict mapping label keys to field names; the default in this case is to directly use the keys of the imported label dictionaries as field names
tags (None) – an optional tag or iterable of tags to attach to each sample
expand_schema (True) – whether to dynamically add new sample fields encountered to the dataset schema. If False, an error is raised if a sample’s schema is not a subset of the dataset schema
dynamic (False) – whether to declare dynamic attributes of embedded document fields that are encountered
add_info (True) – whether to add dataset info from the importer (if any) to the dataset’s info
cleanup (True) – whether to delete the archive after extracting it
progress (None) – whether to render a progress bar (True/False), use the default value fiftyone.config.show_progress_bars (None), or a progress callback function to invoke instead
**kwargs – optional keyword arguments to pass to the constructor of the fiftyone.utils.data.importers.DatasetImporter for the specified dataset_type

Returns

a list of IDs of the samples that were added to the dataset

merge_archive(archive_path, dataset_type=None, data_path=None, labels_path=None, label_field=None, tags=None, key_field='filepath', key_fcn=None, skip_existing=False, insert_new=True, fields=None, omit_fields=None, merge_lists=True, overwrite=True, expand_schema=True, dynamic=False, add_info=True, cleanup=True, progress=None, **kwargs)¶

Merges the contents of the given archive into the dataset.

Note

This method requires the ability to create unique indexes on the key_field of each collection.

See add_archive() if you want to add samples without a uniqueness constraint.

If a directory with the same root name as archive_path exists, it is assumed that this directory contains the extracted contents of the archive, and thus the archive is not re-extracted.

See this guide for example usages of this method and descriptions of the available dataset types.

Note

The following archive formats are explicitly supported:

.zip, .tar, .tar.gz, .tgz, .tar.bz, .tbz

If an archive not in the above list is found, extraction will be attempted via the patool package, which supports many formats but may require that additional system packages be installed.

By default, samples with the same absolute filepath are merged, but you can customize this behavior via the key_field and key_fcn parameters. For example, you could set key_fcn = lambda sample: os.path.basename(sample.filepath) to merge samples with the same base filename.

The behavior of this method is highly customizable. By default, all top-level fields from the imported samples are merged in, overwriting any existing values for those fields, with the exception of list fields (e.g., tags) and label list fields (e.g., fiftyone.core.labels.Detections fields), in which case the elements of the lists themselves are merged. In the case of label list fields, labels with the same id in both collections are updated rather than duplicated.

To avoid confusion between missing fields and fields whose value is None, None-valued fields are always treated as missing while merging.

This method can be configured in numerous ways, including:

Whether existing samples should be modified or skipped
Whether new samples should be added or omitted
Whether new fields can be added to the dataset schema
Whether list fields should be treated as ordinary fields and merged as a whole rather than merging their elements
Whether to merge only specific fields, or all but certain fields
Mapping input fields to different field names of this dataset

Parameters

archive_path – the path to an archive of a dataset directory
dataset_type (None) – the fiftyone.types.Dataset type of the dataset in archive_path
data_path (None) –
an optional parameter that enables explicit control over the location of the media for certain dataset types. Can be any of the following:
- a folder name like "data" or "data/" specifying a subfolder of dataset_dir in which the media lies
- an absolute directory path in which the media lies. In this case, the archive_path has no effect on the location of the data
- a filename like "data.json" specifying the filename of a JSON manifest file in archive_path that maps UUIDs to media filepaths. Files of this format are generated when passing the export_media="manifest" option to fiftyone.core.collections.SampleCollection.export()
- an absolute filepath to a JSON manifest file. In this case, archive_path has no effect on the location of the data
- a dict mapping filenames to absolute filepaths
By default, it is assumed that the data can be located in the default location within archive_path for the dataset type
labels_path (None) –
an optional parameter that enables explicit control over the location of the labels. Only applicable when importing certain labeled dataset formats. Can be any of the following:
- a type-specific folder name like "labels" or "labels/" or a filename like "labels.json" or "labels.xml" specifying the location in archive_path of the labels file(s)
- an absolute directory or filepath containing the labels file(s). In this case, archive_path has no effect on the location of the labels
For labeled datasets, this parameter defaults to the location in archive_path of the labels for the default layout of the dataset type being imported
label_field (None) – controls the field(s) in which imported labels are stored. Only applicable if dataset_importer is a fiftyone.utils.data.importers.LabeledImageDatasetImporter or fiftyone.utils.data.importers.LabeledVideoDatasetImporter. If the importer produces a single fiftyone.core.labels.Label instance per sample/frame, this argument specifies the name of the field to use; the default is "ground_truth". If the importer produces a dictionary of labels per sample, this argument can be either a string prefix to prepend to each label key or a dict mapping label keys to field names; the default in this case is to directly use the keys of the imported label dictionaries as field names
tags (None) – an optional tag or iterable of tags to attach to each sample
key_field ("filepath") – the sample field to use to decide whether to join with an existing sample
key_fcn (None) – a function that accepts a fiftyone.core.sample.Sample instance and computes a key to decide if two samples should be merged. If a key_fcn is provided, key_field is ignored
skip_existing (False) – whether to skip existing samples (True) or merge them (False)
insert_new (True) – whether to insert new samples (True) or skip them (False)
fields (None) – an optional field or iterable of fields to which to restrict the merge. If provided, fields other than these are omitted from samples when merging or adding samples. One exception is that filepath is always included when adding new samples, since the field is required. This can also be a dict mapping field names of the input collection to field names of this dataset
omit_fields (None) – an optional field or iterable of fields to exclude from the merge. If provided, these fields are omitted from imported samples, if present. One exception is that filepath is always included when adding new samples, since the field is required
merge_lists (True) – whether to merge the elements of list fields (e.g., tags) and label list fields (e.g., fiftyone.core.labels.Detections fields) rather than merging the entire top-level field like other field types. For label lists fields, existing fiftyone.core.label.Label elements are either replaced (when overwrite is True) or kept (when overwrite is False) when their id matches a label from the provided samples
overwrite (True) – whether to overwrite (True) or skip (False) existing fields and label elements
expand_schema (True) – whether to dynamically add new fields encountered to the dataset schema. If False, an error is raised if a sample’s schema is not a subset of the dataset schema
dynamic (False) – whether to declare dynamic attributes of embedded document fields that are encountered
add_info (True) – whether to add dataset info from the importer (if any) to the dataset
cleanup (True) – whether to delete the archive after extracting it
progress (None) – whether to render a progress bar (True/False), use the default value fiftyone.config.show_progress_bars (None), or a progress callback function to invoke instead
**kwargs – optional keyword arguments to pass to the constructor of the fiftyone.utils.data.importers.DatasetImporter for the specified dataset_type

add_importer(dataset_importer, label_field=None, tags=None, expand_schema=True, dynamic=False, add_info=True, progress=None)¶

Adds the samples from the given fiftyone.utils.data.importers.DatasetImporter to the dataset.

See this guide for more details about importing datasets in custom formats by defining your own DatasetImporter.

Parameters

dataset_importer – a fiftyone.utils.data.importers.DatasetImporter
label_field (None) – controls the field(s) in which imported labels are stored. Only applicable if dataset_importer is a fiftyone.utils.data.importers.LabeledImageDatasetImporter or fiftyone.utils.data.importers.LabeledVideoDatasetImporter. If the importer produces a single fiftyone.core.labels.Label instance per sample/frame, this argument specifies the name of the field to use; the default is "ground_truth". If the importer produces a dictionary of labels per sample, this argument can be either a string prefix to prepend to each label key or a dict mapping label keys to field names; the default in this case is to directly use the keys of the imported label dictionaries as field names
tags (None) – an optional tag or iterable of tags to attach to each sample
expand_schema (True) – whether to dynamically add new sample fields encountered to the dataset schema. If False, an error is raised if a sample’s schema is not a subset of the dataset schema
dynamic (False) – whether to declare dynamic attributes of embedded document fields that are encountered
add_info (True) – whether to add dataset info from the importer (if any) to the dataset’s info
progress (None) – whether to render a progress bar (True/False), use the default value fiftyone.config.show_progress_bars (None), or a progress callback function to invoke instead

Returns

a list of IDs of the samples that were added to the dataset

merge_importer(dataset_importer, label_field=None, tags=None, key_field='filepath', key_fcn=None, skip_existing=False, insert_new=True, fields=None, omit_fields=None, merge_lists=True, overwrite=True, expand_schema=True, dynamic=False, add_info=True, progress=None)¶

Merges the samples from the given fiftyone.utils.data.importers.DatasetImporter into the dataset.

Note

This method requires the ability to create unique indexes on the key_field of each collection.

See add_importer() if you want to add samples without a uniqueness constraint.

See this guide for more details about importing datasets in custom formats by defining your own DatasetImporter.

By default, samples with the same absolute filepath are merged, but you can customize this behavior via the key_field and key_fcn parameters. For example, you could set key_fcn = lambda sample: os.path.basename(sample.filepath) to merge samples with the same base filename.

The behavior of this method is highly customizable. By default, all top-level fields from the imported samples are merged in, overwriting any existing values for those fields, with the exception of list fields (e.g., tags) and label list fields (e.g., fiftyone.core.labels.Detections fields), in which case the elements of the lists themselves are merged. In the case of label list fields, labels with the same id in both collections are updated rather than duplicated.

To avoid confusion between missing fields and fields whose value is None, None-valued fields are always treated as missing while merging.

This method can be configured in numerous ways, including:

Whether existing samples should be modified or skipped
Whether new samples should be added or omitted
Whether new fields can be added to the dataset schema
Whether list fields should be treated as ordinary fields and merged as a whole rather than merging their elements
Whether to merge only specific fields, or all but certain fields
Mapping input fields to different field names of this dataset

Parameters

dataset_importer – a fiftyone.utils.data.importers.DatasetImporter
label_field (None) – controls the field(s) in which imported labels are stored. Only applicable if dataset_importer is a fiftyone.utils.data.importers.LabeledImageDatasetImporter or fiftyone.utils.data.importers.LabeledVideoDatasetImporter. If the importer produces a single fiftyone.core.labels.Label instance per sample/frame, this argument specifies the name of the field to use; the default is "ground_truth". If the importer produces a dictionary of labels per sample, this argument can be either a string prefix to prepend to each label key or a dict mapping label keys to field names; the default in this case is to directly use the keys of the imported label dictionaries as field names
tags (None) – an optional tag or iterable of tags to attach to each sample
key_field ("filepath") – the sample field to use to decide whether to join with an existing sample
key_fcn (None) – a function that accepts a fiftyone.core.sample.Sample instance and computes a key to decide if two samples should be merged. If a key_fcn is provided, key_field is ignored
skip_existing (False) – whether to skip existing samples (True) or merge them (False)
insert_new (True) – whether to insert new samples (True) or skip them (False)
fields (None) – an optional field or iterable of fields to which to restrict the merge. If provided, fields other than these are omitted from samples when merging or adding samples. One exception is that filepath is always included when adding new samples, since the field is required. This can also be a dict mapping field names of the input collection to field names of this dataset
omit_fields (None) – an optional field or iterable of fields to exclude from the merge. If provided, these fields are omitted from imported samples, if present. One exception is that filepath is always included when adding new samples, since the field is required
merge_lists (True) – whether to merge the elements of list fields (e.g., tags) and label list fields (e.g., fiftyone.core.labels.Detections fields) rather than merging the entire top-level field like other field types. For label lists fields, existing fiftyone.core.label.Label elements are either replaced (when overwrite is True) or kept (when overwrite is False) when their id matches a label from the provided samples
overwrite (True) – whether to overwrite (True) or skip (False) existing fields and label elements
expand_schema (True) – whether to dynamically add new fields encountered to the dataset schema. If False, an error is raised if a sample’s schema is not a subset of the dataset schema
dynamic (False) – whether to declare dynamic attributes of embedded document fields that are encountered
add_info (True) – whether to add dataset info from the importer (if any) to the dataset
progress (None) – whether to render a progress bar (True/False), use the default value fiftyone.config.show_progress_bars (None), or a progress callback function to invoke instead

add_images(paths_or_samples, sample_parser=None, tags=None, progress=None)¶

Adds the given images to the dataset.

This operation does not read the images.

See this guide for more details about adding images to a dataset by defining your own UnlabeledImageSampleParser.

Parameters

paths_or_samples – an iterable of data. If no sample_parser is provided, this must be an iterable of image paths. If a sample_parser is provided, this can be an arbitrary iterable whose elements can be parsed by the sample parser
sample_parser (None) – a fiftyone.utils.data.parsers.UnlabeledImageSampleParser instance to use to parse the samples
tags (None) – an optional tag or iterable of tags to attach to each sample
progress (None) – whether to render a progress bar (True/False), use the default value fiftyone.config.show_progress_bars (None), or a progress callback function to invoke instead

Returns

a list of IDs of the samples that were added to the dataset

add_labeled_images(samples, sample_parser, label_field=None, tags=None, expand_schema=True, dynamic=False, progress=None)¶

Adds the given labeled images to the dataset.

This operation will iterate over all provided samples, but the images will not be read (unless the sample parser requires it in order to compute image metadata).

See this guide for more details about adding labeled images to a dataset by defining your own LabeledImageSampleParser.

Parameters

samples – an iterable of data
sample_parser – a fiftyone.utils.data.parsers.LabeledImageSampleParser instance to use to parse the samples
label_field (None) – controls the field(s) in which imported labels are stored. If the parser produces a single fiftyone.core.labels.Label instance per sample, this argument specifies the name of the field to use; the default is "ground_truth". If the parser produces a dictionary of labels per sample, this argument can be either a string prefix to prepend to each label key or a dict mapping label keys to field names; the default in this case is to directly use the keys of the imported label dictionaries as field names
tags (None) – an optional tag or iterable of tags to attach to each sample
expand_schema (True) – whether to dynamically add new sample fields encountered to the dataset schema. If False, an error is raised if a sample’s schema is not a subset of the dataset schema
dynamic (False) – whether to declare dynamic attributes of embedded document fields that are encountered
progress (None) – whether to render a progress bar (True/False), use the default value fiftyone.config.show_progress_bars (None), or a progress callback function to invoke instead

Returns

a list of IDs of the samples that were added to the dataset

add_images_dir(images_dir, tags=None, recursive=True, progress=None)¶

Adds the given directory of images to the dataset.

See fiftyone.types.ImageDirectory for format details. In particular, note that files with non-image MIME types are omitted.

This operation does not read the images.

Parameters

images_dir – a directory of images
tags (None) – an optional tag or iterable of tags to attach to each sample
recursive (True) – whether to recursively traverse subdirectories
progress (None) – whether to render a progress bar (True/False), use the default value fiftyone.config.show_progress_bars (None), or a progress callback function to invoke instead

Returns

a list of IDs of the samples in the dataset

add_images_patt(images_patt, tags=None, progress=None)¶

Adds the given glob pattern of images to the dataset.

This operation does not read the images.

Parameters

images_patt – a glob pattern of images like /path/to/images/*.jpg
tags (None) – an optional tag or iterable of tags to attach to each sample
progress (None) – whether to render a progress bar (True/False), use the default value fiftyone.config.show_progress_bars (None), or a progress callback function to invoke instead

Returns

a list of IDs of the samples in the dataset

ingest_images(paths_or_samples, sample_parser=None, tags=None, dataset_dir=None, image_format=None, progress=None)¶

Ingests the given iterable of images into the dataset.

The images are read in-memory and written to dataset_dir.

See this guide for more details about ingesting images into a dataset by defining your own UnlabeledImageSampleParser.

Parameters

paths_or_samples – an iterable of data. If no sample_parser is provided, this must be an iterable of image paths. If a sample_parser is provided, this can be an arbitrary iterable whose elements can be parsed by the sample parser
sample_parser (None) – a fiftyone.utils.data.parsers.UnlabeledImageSampleParser instance to use to parse the samples
tags (None) – an optional tag or iterable of tags to attach to each sample
dataset_dir (None) – the directory in which the images will be written. By default, get_default_dataset_dir() is used
image_format (None) – the image format to use to write the images to disk. By default, fiftyone.config.default_image_ext is used
progress (None) – whether to render a progress bar (True/False), use the default value fiftyone.config.show_progress_bars (None), or a progress callback function to invoke instead

Returns

a list of IDs of the samples in the dataset

ingest_labeled_images(samples, sample_parser, label_field=None, tags=None, expand_schema=True, dynamic=False, dataset_dir=None, image_format=None, progress=None)¶

Ingests the given iterable of labeled image samples into the dataset.

The images are read in-memory and written to dataset_dir.

See this guide for more details about ingesting labeled images into a dataset by defining your own LabeledImageSampleParser.

Parameters

samples – an iterable of data
sample_parser – a fiftyone.utils.data.parsers.LabeledImageSampleParser instance to use to parse the samples
label_field (None) – controls the field(s) in which imported labels are stored. If the parser produces a single fiftyone.core.labels.Label instance per sample, this argument specifies the name of the field to use; the default is "ground_truth". If the parser produces a dictionary of labels per sample, this argument can be either a string prefix to prepend to each label key or a dict mapping label keys to field names; the default in this case is to directly use the keys of the imported label dictionaries as field names
tags (None) – an optional tag or iterable of tags to attach to each sample
expand_schema (True) – whether to dynamically add new sample fields encountered to the dataset schema. If False, an error is raised if the sample’s schema is not a subset of the dataset schema
dynamic (False) – whether to declare dynamic attributes of embedded document fields that are encountered
dataset_dir (None) – the directory in which the images will be written. By default, get_default_dataset_dir() is used
image_format (None) – the image format to use to write the images to disk. By default, fiftyone.config.default_image_ext is used
progress (None) – whether to render a progress bar (True/False), use the default value fiftyone.config.show_progress_bars (None), or a progress callback function to invoke instead

Returns

a list of IDs of the samples in the dataset

add_videos(paths_or_samples, sample_parser=None, tags=None, progress=None)¶

Adds the given videos to the dataset.

This operation does not read the videos.

See this guide for more details about adding videos to a dataset by defining your own UnlabeledVideoSampleParser.

Parameters

paths_or_samples – an iterable of data. If no sample_parser is provided, this must be an iterable of video paths. If a sample_parser is provided, this can be an arbitrary iterable whose elements can be parsed by the sample parser
sample_parser (None) – a fiftyone.utils.data.parsers.UnlabeledVideoSampleParser instance to use to parse the samples
tags (None) – an optional tag or iterable of tags to attach to each sample
progress (None) – whether to render a progress bar (True/False), use the default value fiftyone.config.show_progress_bars (None), or a progress callback function to invoke instead

Returns

a list of IDs of the samples that were added to the dataset

add_labeled_videos(samples, sample_parser, label_field=None, tags=None, expand_schema=True, dynamic=False, progress=None)¶

Adds the given labeled videos to the dataset.

This operation will iterate over all provided samples, but the videos will not be read/decoded/etc.

See this guide for more details about adding labeled videos to a dataset by defining your own LabeledVideoSampleParser.

Parameters

samples – an iterable of data
sample_parser – a fiftyone.utils.data.parsers.LabeledVideoSampleParser instance to use to parse the samples
label_field ("ground_truth") – controls the field(s) in which imported labels are stored. If the parser produces a single fiftyone.core.labels.Label instance per sample/frame, this argument specifies the name of the field to use; the default is "ground_truth". If the parser produces a dictionary of labels per sample/frame, this argument can be either a string prefix to prepend to each label key or a dict mapping label keys to field names; the default in this case is to directly use the keys of the imported label dictionaries as field names
label_field – the name (or root name) of the frame field(s) to use for the labels
tags (None) – an optional tag or iterable of tags to attach to each sample
expand_schema (True) – whether to dynamically add new sample fields encountered to the dataset schema. If False, an error is raised if a sample’s schema is not a subset of the dataset schema
dynamic (False) – whether to declare dynamic attributes of embedded document fields that are encountered
progress (None) – whether to render a progress bar (True/False), use the default value fiftyone.config.show_progress_bars (None), or a progress callback function to invoke instead

Returns

a list of IDs of the samples that were added to the dataset

add_videos_dir(videos_dir, tags=None, recursive=True, progress=None)¶

Adds the given directory of videos to the dataset.

See fiftyone.types.VideoDirectory for format details. In particular, note that files with non-video MIME types are omitted.

This operation does not read/decode the videos.

Parameters

videos_dir – a directory of videos
tags (None) – an optional tag or iterable of tags to attach to each sample
recursive (True) – whether to recursively traverse subdirectories
progress (None) – whether to render a progress bar (True/False), use the default value fiftyone.config.show_progress_bars (None), or a progress callback function to invoke instead

Returns

a list of IDs of the samples in the dataset

add_videos_patt(videos_patt, tags=None, progress=None)¶

Adds the given glob pattern of videos to the dataset.

This operation does not read/decode the videos.

Parameters

videos_patt – a glob pattern of videos like /path/to/videos/*.mp4
tags (None) – an optional tag or iterable of tags to attach to each sample
progress (None) – whether to render a progress bar (True/False), use the default value fiftyone.config.show_progress_bars (None), or a progress callback function to invoke instead

Returns

a list of IDs of the samples in the dataset

ingest_videos(paths_or_samples, sample_parser=None, tags=None, dataset_dir=None, progress=None)¶

Ingests the given iterable of videos into the dataset.

The videos are copied to dataset_dir.

See this guide for more details about ingesting videos into a dataset by defining your own UnlabeledVideoSampleParser.

Parameters

paths_or_samples – an iterable of data. If no sample_parser is provided, this must be an iterable of video paths. If a sample_parser is provided, this can be an arbitrary iterable whose elements can be parsed by the sample parser
sample_parser (None) – a fiftyone.utils.data.parsers.UnlabeledVideoSampleParser instance to use to parse the samples
tags (None) – an optional tag or iterable of tags to attach to each sample
dataset_dir (None) – the directory in which the videos will be written. By default, get_default_dataset_dir() is used
progress (None) – whether to render a progress bar (True/False), use the default value fiftyone.config.show_progress_bars (None), or a progress callback function to invoke instead

Returns

a list of IDs of the samples in the dataset

ingest_labeled_videos(samples, sample_parser, tags=None, expand_schema=True, dynamic=False, dataset_dir=None, progress=None)¶

Ingests the given iterable of labeled video samples into the dataset.

The videos are copied to dataset_dir.

See this guide for more details about ingesting labeled videos into a dataset by defining your own LabeledVideoSampleParser.

Parameters

samples – an iterable of data
sample_parser – a fiftyone.utils.data.parsers.LabeledVideoSampleParser instance to use to parse the samples
tags (None) – an optional tag or iterable of tags to attach to each sample
expand_schema (True) – whether to dynamically add new sample fields encountered to the dataset schema. If False, an error is raised if the sample’s schema is not a subset of the dataset schema
dynamic (False) – whether to declare dynamic attributes of embedded document fields that are encountered
dataset_dir (None) – the directory in which the videos will be written. By default, get_default_dataset_dir() is used
progress (None) – whether to render a progress bar (True/False), use the default value fiftyone.config.show_progress_bars (None), or a progress callback function to invoke instead

Returns

a list of IDs of the samples in the dataset

classmethod from_dir(dataset_dir=None, dataset_type=None, data_path=None, labels_path=None, name=None, persistent=False, overwrite=False, label_field=None, tags=None, dynamic=False, progress=None, **kwargs)¶

Creates a Dataset from the contents of the given directory.

You can create datasets with this method via the following basic patterns:

Provide dataset_dir and dataset_type to import the contents of a directory that is organized in the default layout for the dataset type as documented in this guide
Provide dataset_type along with data_path, labels_path, or other type-specific parameters to perform a customized import. This syntax provides the flexibility to, for example, perform labels-only imports or imports where the source media lies in a different location than the labels

In either workflow, the remaining parameters of this method can be provided to further configure the import.

See this guide for example usages of this method and descriptions of the available dataset types.

Parameters

dataset_dir (None) – the dataset directory. This can be omitted if you provide arguments such as data_path and labels_path
dataset_type (None) – the fiftyone.types.Dataset type of the dataset
data_path (None) –
an optional parameter that enables explicit control over the location of the media for certain dataset types. Can be any of the following:
- a folder name like "data" or "data/" specifying a subfolder of dataset_dir in which the media lies
- an absolute directory path in which the media lies. In this case, the dataset_dir has no effect on the location of the data
- a filename like "data.json" specifying the filename of a JSON manifest file in dataset_dir that maps UUIDs to media filepaths. Files of this format are generated when passing the export_media="manifest" option to fiftyone.core.collections.SampleCollection.export()
- an absolute filepath to a JSON manifest file. In this case, dataset_dir has no effect on the location of the data
- a dict mapping filenames to absolute filepaths
By default, it is assumed that the data can be located in the default location within dataset_dir for the dataset type
labels_path (None) –
an optional parameter that enables explicit control over the location of the labels. Only applicable when importing certain labeled dataset formats. Can be any of the following:
- a type-specific folder name like "labels" or "labels/" or a filename like "labels.json" or "labels.xml" specifying the location in dataset_dir of the labels file(s)
- an absolute directory or filepath containing the labels file(s). In this case, dataset_dir has no effect on the location of the labels
For labeled datasets, this parameter defaults to the location in dataset_dir of the labels for the default layout of the dataset type being imported
name (None) – a name for the dataset. By default, get_default_dataset_name() is used
persistent (False) – whether the dataset should persist in the database after the session terminates
overwrite (False) – whether to overwrite an existing dataset of the same name
label_field (None) – controls the field(s) in which imported labels are stored. Only applicable if dataset_importer is a fiftyone.utils.data.importers.LabeledImageDatasetImporter or fiftyone.utils.data.importers.LabeledVideoDatasetImporter. If the importer produces a single fiftyone.core.labels.Label instance per sample/frame, this argument specifies the name of the field to use; the default is "ground_truth". If the importer produces a dictionary of labels per sample, this argument can be either a string prefix to prepend to each label key or a dict mapping label keys to field names; the default in this case is to directly use the keys of the imported label dictionaries as field names
tags (None) – an optional tag or iterable of tags to attach to each sample
dynamic (False) – whether to declare dynamic attributes of embedded document fields that are encountered
progress (None) – whether to render a progress bar (True/False), use the default value fiftyone.config.show_progress_bars (None), or a progress callback function to invoke instead
**kwargs – optional keyword arguments to pass to the constructor of the fiftyone.utils.data.importers.DatasetImporter for the specified dataset_type

Returns

a Dataset

classmethod from_archive(archive_path, dataset_type=None, data_path=None, labels_path=None, name=None, persistent=False, overwrite=False, label_field=None, tags=None, dynamic=False, cleanup=True, progress=None, **kwargs)¶

Creates a Dataset from the contents of the given archive.

If a directory with the same root name as archive_path exists, it is assumed that this directory contains the extracted contents of the archive, and thus the archive is not re-extracted.

See this guide for example usages of this method and descriptions of the available dataset types.

Note

The following archive formats are explicitly supported:

.zip, .tar, .tar.gz, .tgz, .tar.bz, .tbz

If an archive not in the above list is found, extraction will be attempted via the patool package, which supports many formats but may require that additional system packages be installed.

Parameters

archive_path – the path to an archive of a dataset directory
dataset_type (None) – the fiftyone.types.Dataset type of the dataset in archive_path
data_path (None) –
an optional parameter that enables explicit control over the location of the media for certain dataset types. Can be any of the following:
- a folder name like "data" or "data/" specifying a subfolder of dataset_dir in which the media lies
- an absolute directory path in which the media lies. In this case, the archive_path has no effect on the location of the data
- a filename like "data.json" specifying the filename of a JSON manifest file in archive_path that maps UUIDs to media filepaths. Files of this format are generated when passing the export_media="manifest" option to fiftyone.core.collections.SampleCollection.export()
- an absolute filepath to a JSON manifest file. In this case, archive_path has no effect on the location of the data
- a dict mapping filenames to absolute filepaths
By default, it is assumed that the data can be located in the default location within archive_path for the dataset type
labels_path (None) –
an optional parameter that enables explicit control over the location of the labels. Only applicable when importing certain labeled dataset formats. Can be any of the following:
- a type-specific folder name like "labels" or "labels/" or a filename like "labels.json" or "labels.xml" specifying the location in archive_path of the labels file(s)
- an absolute directory or filepath containing the labels file(s). In this case, archive_path has no effect on the location of the labels
For labeled datasets, this parameter defaults to the location in archive_path of the labels for the default layout of the dataset type being imported
name (None) – a name for the dataset. By default, get_default_dataset_name() is used
persistent (False) – whether the dataset should persist in the database after the session terminates
overwrite (False) – whether to overwrite an existing dataset of the same name
label_field (None) – controls the field(s) in which imported labels are stored. Only applicable if dataset_importer is a fiftyone.utils.data.importers.LabeledImageDatasetImporter or fiftyone.utils.data.importers.LabeledVideoDatasetImporter. If the importer produces a single fiftyone.core.labels.Label instance per sample/frame, this argument specifies the name of the field to use; the default is "ground_truth". If the importer produces a dictionary of labels per sample, this argument can be either a string prefix to prepend to each label key or a dict mapping label keys to field names; the default in this case is to directly use the keys of the imported label dictionaries as field names
tags (None) – an optional tag or iterable of tags to attach to each sample
dynamic (False) – whether to declare dynamic attributes of embedded document fields that are encountered
cleanup (True) – whether to delete the archive after extracting it
progress (None) – whether to render a progress bar (True/False), use the default value fiftyone.config.show_progress_bars (None), or a progress callback function to invoke instead
**kwargs – optional keyword arguments to pass to the constructor of the fiftyone.utils.data.importers.DatasetImporter for the specified dataset_type

Returns

a Dataset

classmethod from_importer(dataset_importer, name=None, persistent=False, overwrite=False, label_field=None, tags=None, dynamic=False, progress=None)¶

Creates a Dataset by importing the samples in the given fiftyone.utils.data.importers.DatasetImporter.

See this guide for more details about providing a custom DatasetImporter to import datasets into FiftyOne.

Parameters

dataset_importer – a fiftyone.utils.data.importers.DatasetImporter
name (None) – a name for the dataset. By default, get_default_dataset_name() is used
persistent (False) – whether the dataset should persist in the database after the session terminates
overwrite (False) – whether to overwrite an existing dataset of the same name
label_field (None) – controls the field(s) in which imported labels are stored. Only applicable if dataset_importer is a fiftyone.utils.data.importers.LabeledImageDatasetImporter or fiftyone.utils.data.importers.LabeledVideoDatasetImporter. If the importer produces a single fiftyone.core.labels.Label instance per sample/frame, this argument specifies the name of the field to use; the default is "ground_truth". If the importer produces a dictionary of labels per sample, this argument can be either a string prefix to prepend to each label key or a dict mapping label keys to field names; the default in this case is to directly use the keys of the imported label dictionaries as field names
tags (None) – an optional tag or iterable of tags to attach to each sample
dynamic (False) – whether to declare dynamic attributes of embedded document fields that are encountered
progress (None) – whether to render a progress bar (True/False), use the default value fiftyone.config.show_progress_bars (None), or a progress callback function to invoke instead

Returns

a Dataset

classmethod from_images(paths_or_samples, sample_parser=None, name=None, persistent=False, overwrite=False, tags=None, progress=None)¶

Creates a Dataset from the given images.

This operation does not read the images.

See this guide for more details about providing a custom UnlabeledImageSampleParser to load image samples into FiftyOne.

Parameters

paths_or_samples – an iterable of data. If no sample_parser is provided, this must be an iterable of image paths. If a sample_parser is provided, this can be an arbitrary iterable whose elements can be parsed by the sample parser
sample_parser (None) – a fiftyone.utils.data.parsers.UnlabeledImageSampleParser instance to use to parse the samples
name (None) – a name for the dataset. By default, get_default_dataset_name() is used
persistent (False) – whether the dataset should persist in the database after the session terminates
overwrite (False) – whether to overwrite an existing dataset of the same name
tags (None) – an optional tag or iterable of tags to attach to each sample
progress (None) – whether to render a progress bar (True/False), use the default value fiftyone.config.show_progress_bars (None), or a progress callback function to invoke instead

Returns

a Dataset

classmethod from_labeled_images(samples, sample_parser, name=None, persistent=False, overwrite=False, label_field=None, tags=None, dynamic=False, progress=None)¶

Creates a Dataset from the given labeled images.

This operation will iterate over all provided samples, but the images will not be read.

See this guide for more details about providing a custom LabeledImageSampleParser to load labeled image samples into FiftyOne.

Parameters

samples – an iterable of data
sample_parser – a fiftyone.utils.data.parsers.LabeledImageSampleParser instance to use to parse the samples
name (None) – a name for the dataset. By default, get_default_dataset_name() is used
persistent (False) – whether the dataset should persist in the database after the session terminates
overwrite (False) – whether to overwrite an existing dataset of the same name
label_field (None) – controls the field(s) in which imported labels are stored. If the parser produces a single fiftyone.core.labels.Label instance per sample, this argument specifies the name of the field to use; the default is "ground_truth". If the parser produces a dictionary of labels per sample, this argument can be either a string prefix to prepend to each label key or a dict mapping label keys to field names; the default in this case is to directly use the keys of the imported label dictionaries as field names
tags (None) – an optional tag or iterable of tags to attach to each sample
dynamic (False) – whether to declare dynamic attributes of embedded document fields that are encountered
progress (None) – whether to render a progress bar (True/False), use the default value fiftyone.config.show_progress_bars (None), or a progress callback function to invoke instead

Returns

a Dataset

classmethod from_images_dir(images_dir, name=None, persistent=False, overwrite=False, tags=None, recursive=True, progress=None)¶

Creates a Dataset from the given directory of images.

This operation does not read the images.

Parameters

images_dir – a directory of images
name (None) – a name for the dataset. By default, get_default_dataset_name() is used
persistent (False) – whether the dataset should persist in the database after the session terminates
overwrite (False) – whether to overwrite an existing dataset of the same name
tags (None) – an optional tag or iterable of tags to attach to each sample
recursive (True) – whether to recursively traverse subdirectories
progress (None) – whether to render a progress bar (True/False), use the default value fiftyone.config.show_progress_bars (None), or a progress callback function to invoke instead

Returns

a Dataset

classmethod from_images_patt(images_patt, name=None, persistent=False, overwrite=False, tags=None, progress=None)¶

Creates a Dataset from the given glob pattern of images.

This operation does not read the images.

Parameters

images_patt – a glob pattern of images like /path/to/images/*.jpg
name (None) – a name for the dataset. By default, get_default_dataset_name() is used
persistent (False) – whether the dataset should persist in the database after the session terminates
overwrite (False) – whether to overwrite an existing dataset of the same name
tags (None) – an optional tag or iterable of tags to attach to each sample
progress (None) – whether to render a progress bar (True/False), use the default value fiftyone.config.show_progress_bars (None), or a progress callback function to invoke instead

Returns

a Dataset

classmethod from_videos(paths_or_samples, sample_parser=None, name=None, persistent=False, overwrite=False, tags=None, progress=None)¶

Creates a Dataset from the given videos.

This operation does not read/decode the videos.

See this guide for more details about providing a custom UnlabeledVideoSampleParser to load video samples into FiftyOne.

Parameters

paths_or_samples – an iterable of data. If no sample_parser is provided, this must be an iterable of video paths. If a sample_parser is provided, this can be an arbitrary iterable whose elements can be parsed by the sample parser
sample_parser (None) – a fiftyone.utils.data.parsers.UnlabeledVideoSampleParser instance to use to parse the samples
name (None) – a name for the dataset. By default, get_default_dataset_name() is used
persistent (False) – whether the dataset should persist in the database after the session terminates
overwrite (False) – whether to overwrite an existing dataset of the same name
tags (None) – an optional tag or iterable of tags to attach to each sample
progress (None) – whether to render a progress bar (True/False), use the default value fiftyone.config.show_progress_bars (None), or a progress callback function to invoke instead

Returns

a Dataset

classmethod from_labeled_videos(samples, sample_parser, name=None, persistent=False, overwrite=False, label_field=None, tags=None, dynamic=False, progress=None)¶

Creates a Dataset from the given labeled videos.

This operation will iterate over all provided samples, but the videos will not be read/decoded/etc.

See this guide for more details about providing a custom LabeledVideoSampleParser to load labeled video samples into FiftyOne.

Parameters

samples – an iterable of data
sample_parser – a fiftyone.utils.data.parsers.LabeledVideoSampleParser instance to use to parse the samples
name (None) – a name for the dataset. By default, get_default_dataset_name() is used
persistent (False) – whether the dataset should persist in the database after the session terminates
overwrite (False) – whether to overwrite an existing dataset of the same name
label_field (None) – controls the field(s) in which imported labels are stored. If the parser produces a single fiftyone.core.labels.Label instance per sample/frame, this argument specifies the name of the field to use; the default is "ground_truth". If the parser produces a dictionary of labels per sample/frame, this argument can be either a string prefix to prepend to each label key or a dict mapping label keys to field names; the default in this case is to directly use the keys of the imported label dictionaries as field names
tags (None) – an optional tag or iterable of tags to attach to each sample
dynamic (False) – whether to declare dynamic attributes of embedded document fields that are encountered
progress (None) – whether to render a progress bar (True/False), use the default value fiftyone.config.show_progress_bars (None), or a progress callback function to invoke instead

Returns

a Dataset

classmethod from_videos_dir(videos_dir, name=None, persistent=False, overwrite=False, tags=None, recursive=True, progress=None)¶

Creates a Dataset from the given directory of videos.

This operation does not read/decode the videos.

Parameters

videos_dir – a directory of videos
name (None) – a name for the dataset. By default, get_default_dataset_name() is used
persistent (False) – whether the dataset should persist in the database after the session terminates
overwrite (False) – whether to overwrite an existing dataset of the same name
tags (None) – an optional tag or iterable of tags to attach to each sample
recursive (True) – whether to recursively traverse subdirectories

Returns

a Dataset

classmethod from_videos_patt(videos_patt, name=None, persistent=False, overwrite=False, tags=None, progress=None)¶

Creates a Dataset from the given glob pattern of videos.

This operation does not read/decode the videos.

Parameters

videos_patt – a glob pattern of videos like /path/to/videos/*.mp4
name (None) – a name for the dataset. By default, get_default_dataset_name() is used
persistent (False) – whether the dataset should persist in the database after the session terminates
overwrite (False) – whether to overwrite an existing dataset of the same name
tags (None) – an optional tag or iterable of tags to attach to each sample

Returns

a Dataset

classmethod from_dict(d, name=None, persistent=False, overwrite=False, rel_dir=None, frame_labels_dir=None, progress=None)¶

Loads a Dataset from a JSON dictionary generated by fiftyone.core.collections.SampleCollection.to_dict().

The JSON dictionary can contain an export of any fiftyone.core.collections.SampleCollection, e.g., Dataset or fiftyone.core.view.DatasetView.

Parameters

d – a JSON dictionary
name (None) – a name for the new dataset
persistent (False) – whether the dataset should persist in the database after the session terminates
overwrite (False) – whether to overwrite an existing dataset of the same name
rel_dir (None) – a relative directory to prepend to the filepath of each sample if the filepath is not absolute (begins with a path separator). The path is converted to an absolute path (if necessary) via fiftyone.core.storage.normalize_path()
frame_labels_dir (None) – a directory of per-sample JSON files containing the frame labels for video samples. If omitted, it is assumed that the frame labels are included directly in the provided JSON dict. Only applicable to datasets that contain videos
progress (None) – whether to render a progress bar (True/False), use the default value fiftyone.config.show_progress_bars (None), or a progress callback function to invoke instead

Returns

a Dataset

classmethod from_json(path_or_str, name=None, persistent=False, overwrite=False, rel_dir=None, frame_labels_dir=None, progress=None)¶

Loads a Dataset from JSON generated by fiftyone.core.collections.SampleCollection.write_json() or fiftyone.core.collections.SampleCollection.to_json().

The JSON file can contain an export of any fiftyone.core.collections.SampleCollection, e.g., Dataset or fiftyone.core.view.DatasetView.

Parameters

path_or_str – the path to a JSON file on disk or a JSON string
name (None) – a name for the new dataset
persistent (False) – whether the dataset should persist in the database after the session terminates
overwrite (False) – whether to overwrite an existing dataset of the same name
rel_dir (None) – a relative directory to prepend to the filepath of each sample, if the filepath is not absolute (begins with a path separator). The path is converted to an absolute path (if necessary) via fiftyone.core.storage.normalize_path()
progress (None) – whether to render a progress bar (True/False), use the default value fiftyone.config.show_progress_bars (None), or a progress callback function to invoke instead

Returns

a Dataset

reload()¶: Reloads the dataset and any in-memory samples from the database.

clear_cache()¶

Clears the dataset’s in-memory cache.

Dataset caches may contain sample/frame singletons and annotation/brain/evaluation/custom runs.

add_stage(stage)¶

Applies the given fiftyone.core.stages.ViewStage to the collection.

Parameters: stage – a fiftyone.core.stages.ViewStage
Returns: a fiftyone.core.view.DatasetView

aggregate(aggregations)¶

Aggregates one or more fiftyone.core.aggregations.Aggregation instances.

Note that it is best practice to group aggregations into a single call to aggregate(), as this will be more efficient than performing multiple aggregations in series.

Parameters: aggregations – an fiftyone.core.aggregations.Aggregation or iterable of fiftyone.core.aggregations.Aggregation instances
Returns: an aggregation result or list of aggregation results corresponding to the input aggregation(s)

annotate(anno_key, label_schema=None, label_field=None, label_type=None, classes=None, attributes=True, mask_targets=None, allow_additions=True, allow_deletions=True, allow_label_edits=True, allow_index_edits=True, allow_spatial_edits=True, media_field='filepath', backend=None, launch_editor=False, **kwargs)¶

Exports the samples and optional label field(s) in this collection to the given annotation backend.

The backend parameter controls which annotation backend to use. Depending on the backend you use, you may want/need to provide extra keyword arguments to this function for the constructor of the backend’s fiftyone.utils.annotations.AnnotationBackendConfig class.

The natively provided backends and their associated config classes are:

"cvat": fiftyone.utils.cvat.CVATBackendConfig
"labelstudio": fiftyone.utils.labelstudio.LabelStudioBackendConfig
"labelbox": fiftyone.utils.labelbox.LabelboxBackendConfig

See this page for more information about using this method, including how to define label schemas and how to configure login credentials for your annotation provider.

Parameters

anno_key – a string key to use to refer to this annotation run
label_schema (None) – a dictionary defining the label schema to use. If this argument is provided, it takes precedence over the other schema-related arguments
label_field (None) – a string indicating a new or existing label field to annotate
label_type (None) –
a string indicating the type of labels to annotate. The possible values are:
- "classification": a single classification stored in fiftyone.core.labels.Classification fields
- "classifications": multilabel classifications stored in fiftyone.core.labels.Classifications fields
- "detections": object detections stored in fiftyone.core.labels.Detections fields
- "instances": instance segmentations stored in fiftyone.core.labels.Detections fields with their mask attributes populated
- "polylines": polylines stored in fiftyone.core.labels.Polylines fields with their filled attributes set to False
- "polygons": polygons stored in fiftyone.core.labels.Polylines fields with their filled attributes set to True
- "keypoints": keypoints stored in fiftyone.core.labels.Keypoints fields
- "segmentation": semantic segmentations stored in fiftyone.core.labels.Segmentation fields
- "scalar": scalar labels stored in fiftyone.core.fields.IntField, fiftyone.core.fields.FloatField, fiftyone.core.fields.StringField, or fiftyone.core.fields.BooleanField fields
All new label fields must have their type specified via this argument or in label_schema. Note that annotation backends may not support all label types
classes (None) – a list of strings indicating the class options for label_field or all fields in label_schema without classes specified. All new label fields must have a class list provided via one of the supported methods. For existing label fields, if classes are not provided by this argument nor label_schema, they are retrieved from get_classes() if possible, or else the observed labels on your dataset are used
attributes (True) –
specifies the label attributes of each label field to include (other than their label, which is always included) in the annotation export. Can be any of the following:
- True: export all label attributes
- False: don’t export any custom label attributes
- a list of label attributes to export
- a dict mapping attribute names to dicts specifying the type, values, and default for each attribute
If a label_schema is also provided, this parameter determines which attributes are included for all fields that do not explicitly define their per-field attributes (in addition to any per-class attributes)
mask_targets (None) – a dict mapping pixel values to semantic label strings. Only applicable when annotating semantic segmentations
allow_additions (True) – whether to allow new labels to be added. Only applicable when editing existing label fields
allow_deletions (True) – whether to allow labels to be deleted. Only applicable when editing existing label fields
allow_label_edits (True) – whether to allow the label attribute of existing labels to be modified. Only applicable when editing existing fields with label attributes
allow_index_edits (True) – whether to allow the index attribute of existing video tracks to be modified. Only applicable when editing existing frame fields with index attributes
allow_spatial_edits (True) – whether to allow edits to the spatial properties (bounding boxes, vertices, keypoints, masks, etc) of labels. Only applicable when editing existing spatial label fields
media_field ("filepath") – the field containing the paths to the media files to upload
backend (None) – the annotation backend to use. The supported values are fiftyone.annotation_config.backends.keys() and the default is fiftyone.annotation_config.default_backend
launch_editor (False) – whether to launch the annotation backend’s editor after uploading the samples
**kwargs – keyword arguments for the fiftyone.utils.annotations.AnnotationBackendConfig

Returns

an fiftyone.utils.annotations.AnnnotationResults

apply_model(model, label_field='predictions', confidence_thresh=None, store_logits=False, batch_size=None, num_workers=None, skip_failures=True, output_dir=None, rel_dir=None, progress=None, **kwargs)¶

Applies the model to the samples in the collection.

This method supports all of the following cases:

Applying an image model to an image collection
Applying an image model to the frames of a video collection
Applying a video model to a video collection

Parameters

model – a fiftyone.core.models.Model, Hugging Face transformers model, Ultralytics model, SuperGradients model, or Lightning Flash model
label_field ("predictions") – the name of the field in which to store the model predictions. When performing inference on video frames, the “frames.” prefix is optional
confidence_thresh (None) – an optional confidence threshold to apply to any applicable labels generated by the model
store_logits (False) – whether to store logits for the model predictions. This is only supported when the provided model has logits, model.has_logits == True
batch_size (None) – an optional batch size to use, if the model supports batching
num_workers (None) – the number of workers for the torch.utils.data.DataLoader to use. Only applicable for Torch-based models
skip_failures (True) – whether to gracefully continue without raising an error if predictions cannot be generated for a sample. Only applicable to fiftyone.core.models.Model instances
output_dir (None) – an optional output directory in which to write segmentation images. Only applicable if the model generates segmentations. If none is provided, the segmentations are stored in the database
rel_dir (None) – an optional relative directory to strip from each input filepath to generate a unique identifier that is joined with output_dir to generate an output path for each segmentation image. This argument allows for populating nested subdirectories in output_dir that match the shape of the input paths. The path is converted to an absolute path (if necessary) via fiftyone.core.storage.normalize_path()
progress (None) – whether to render a progress bar (True/False), use the default value fiftyone.config.show_progress_bars (None), or a progress callback function to invoke instead
**kwargs – optional model-specific keyword arguments passed through to the underlying inference implementation

bounds(field_or_expr, expr=None, safe=False)¶

Computes the bounds of a numeric field of the collection.

None-valued fields are ignored.

This aggregation is typically applied to numeric field types (or lists of such types):

Examples:

import fiftyone as fo
from fiftyone import ViewField as F

dataset = fo.Dataset()
dataset.add_samples(
    [
        fo.Sample(
            filepath="/path/to/image1.png",
            numeric_field=1.0,
            numeric_list_field=[1, 2, 3],
        ),
        fo.Sample(
            filepath="/path/to/image2.png",
            numeric_field=4.0,
            numeric_list_field=[1, 2],
        ),
        fo.Sample(
            filepath="/path/to/image3.png",
            numeric_field=None,
            numeric_list_field=None,
        ),
    ]
)

#
# Compute the bounds of a numeric field
#

bounds = dataset.bounds("numeric_field")
print(bounds)  # (min, max)

#
# Compute the a bounds of a numeric list field
#

bounds = dataset.bounds("numeric_list_field")
print(bounds)  # (min, max)

#
# Compute the bounds of a transformation of a numeric field
#

bounds = dataset.bounds(2 * (F("numeric_field") + 1))
print(bounds)  # (min, max)

Parameters

field_or_expr –
a field name, embedded.field.name, fiftyone.core.expressions.ViewExpression, or MongoDB expression defining the field or expression to aggregate. This can also be a list or tuple of such arguments, in which case a tuple of corresponding aggregation results (each receiving the same additional keyword arguments, if any) will be returned
expr (None) –
a fiftyone.core.expressions.ViewExpression or MongoDB expression to apply to field_or_expr (which must be a field) before aggregating
safe (False) – whether to ignore nan/inf values when dealing with floating point values

Returns

the (min, max) bounds

compute_embeddings(model, embeddings_field=None, batch_size=None, num_workers=None, skip_failures=True, progress=None, **kwargs)¶

Computes embeddings for the samples in the collection using the given model.

This method supports all the following cases:

Using an image model to compute embeddings for an image collection
Using an image model to compute frame embeddings for a video collection
Using a video model to compute embeddings for a video collection

The model must expose embeddings, i.e., fiftyone.core.models.Model.has_embeddings() must return True.

If an embeddings_field is provided, the embeddings are saved to the samples; otherwise, the embeddings are returned in-memory.

Parameters

model – a fiftyone.core.models.Model, Hugging Face Transformers model, Ultralytics model, SuperGradients model, or Lightning Flash model
embeddings_field (None) – the name of a field in which to store the embeddings. When computing video frame embeddings, the “frames.” prefix is optional
batch_size (None) – an optional batch size to use, if the model supports batching
num_workers (None) – the number of workers for the torch.utils.data.DataLoader to use. Only applicable for Torch-based models
skip_failures (True) – whether to gracefully continue without raising an error if embeddings cannot be generated for a sample. Only applicable to fiftyone.core.models.Model instances
progress (None) – whether to render a progress bar (True/False), use the default value fiftyone.config.show_progress_bars (None), or a progress callback function to invoke instead
**kwargs – optional model-specific keyword arguments passed through to the underlying inference implementation

Returns

None, if an embeddings_field is provided
a num_samples x num_dim array of embeddings, when computing embeddings for image/video collections with image/video models, respectively, and no embeddings_field is provided. If skip_failures is True and any errors are detected, a list of length num_samples is returned instead containing all successfully computed embedding vectors along with None entries for samples for which embeddings could not be computed
a dictionary mapping sample IDs to num_frames x num_dim arrays of embeddings, when computing frame embeddings for video collections using an image model. If skip_failures is True and any errors are detected, the values of this dictionary will contain arrays of embeddings for all frames 1, 2, … until the error occurred, or None if no embeddings were computed at all

Return type

one of the following

compute_metadata(overwrite=False, num_workers=None, skip_failures=True, warn_failures=False, progress=None)¶

Populates the metadata field of all samples in the collection.

Any samples with existing metadata are skipped, unless overwrite == True.

Parameters

overwrite (False) – whether to overwrite existing metadata
num_workers (None) – a suggested number of threads to use
skip_failures (True) – whether to gracefully continue without raising an error if metadata cannot be computed for a sample
warn_failures (False) – whether to log a warning if metadata cannot be computed for a sample
progress (None) – whether to render a progress bar (True/False), use the default value fiftyone.config.show_progress_bars (None), or a progress callback function to invoke instead

compute_patch_embeddings(model, patches_field, embeddings_field=None, force_square=False, alpha=None, handle_missing='skip', batch_size=None, num_workers=None, skip_failures=True, progress=None)¶

Computes embeddings for the image patches defined by patches_field of the samples in the collection using the given model.

This method supports all the following cases:

Using an image model to compute patch embeddings for an image collection
Using an image model to compute frame patch embeddings for a video collection

The model must expose embeddings, i.e., fiftyone.core.models.Model.has_embeddings() must return True.

If an embeddings_field is provided, the embeddings are saved to the samples; otherwise, the embeddings are returned in-memory.

Parameters

model – a fiftyone.core.models.Model, Hugging Face Transformers model, Ultralytics model, SuperGradients model, or Lightning Flash model
patches_field – the name of the field defining the image patches in each sample to embed. Must be of type fiftyone.core.labels.Detection, fiftyone.core.labels.Detections, fiftyone.core.labels.Polyline, or fiftyone.core.labels.Polylines. When computing video frame embeddings, the “frames.” prefix is optional
embeddings_field (None) – the name of a label attribute in which to store the embeddings
force_square (False) – whether to minimally manipulate the patch bounding boxes into squares prior to extraction
alpha (None) – an optional expansion/contraction to apply to the patches before extracting them, in [-1, inf). If provided, the length and width of the box are expanded (or contracted, when alpha < 0) by (100 * alpha)%. For example, set alpha = 0.1 to expand the boxes by 10%, and set alpha = -0.1 to contract the boxes by 10%
handle_missing ("skip") –
how to handle images with no patches. Supported values are:
- ”skip”: skip the image and assign its embedding as None
- ”image”: use the whole image as a single patch
- ”error”: raise an error
batch_size (None) – an optional batch size to use, if the model supports batching
num_workers (None) – the number of workers for the torch.utils.data.DataLoader to use. Only applicable for Torch-based models
skip_failures (True) – whether to gracefully continue without raising an error if embeddings cannot be generated for a sample
progress (None) – whether to render a progress bar (True/False), use the default value fiftyone.config.show_progress_bars (None), or a progress callback function to invoke instead

Returns

None, if an embeddings_field is provided
a dict mapping sample IDs to num_patches x num_dim arrays of patch embeddings, when computing patch embeddings for image collections and no embeddings_field is provided. If skip_failures is True and any errors are detected, this dictionary will contain None values for any samples for which embeddings could not be computed
a dict of dicts mapping sample IDs to frame numbers to num_patches x num_dim arrays of patch embeddings, when computing patch embeddings for the frames of video collections and no embeddings_field is provided. If skip_failures is True and any errors are detected, this nested dict will contain missing or None values to indicate uncomputable embeddings

Return type

one of the following

concat(samples)¶

Concatenates the contents of the given SampleCollection to this collection.

Examples:

import fiftyone as fo
import fiftyone.zoo as foz
from fiftyone import ViewField as F

dataset = foz.load_zoo_dataset("quickstart")

#
# Concatenate two views
#

view1 = dataset.match(F("uniqueness") < 0.2)
view2 = dataset.match(F("uniqueness") > 0.7)

view = view1.concat(view2)

print(view1)
print(view2)
print(view)

#
# Concatenate two patches views
#

gt_objects = dataset.to_patches("ground_truth")

patches1 = gt_objects[:50]
patches2 = gt_objects[-50:]
patches = patches1.concat(patches2)

print(patches1)
print(patches2)
print(patches)

Parameters: samples – a SampleCollection whose contents to append to this collection
Returns: a fiftyone.core.view.DatasetView

count(field_or_expr=None, expr=None, safe=False)¶

Counts the number of field values in the collection.

None-valued fields are ignored.

If no field is provided, the samples themselves are counted.

Examples:

import fiftyone as fo
from fiftyone import ViewField as F

dataset = fo.Dataset()
dataset.add_samples(
    [
        fo.Sample(
            filepath="/path/to/image1.png",
            predictions=fo.Detections(
                detections=[
                    fo.Detection(label="cat"),
                    fo.Detection(label="dog"),
                ]
            ),
        ),
        fo.Sample(
            filepath="/path/to/image2.png",
            predictions=fo.Detections(
                detections=[
                    fo.Detection(label="cat"),
                    fo.Detection(label="rabbit"),
                    fo.Detection(label="squirrel"),
                ]
            ),
        ),
        fo.Sample(
            filepath="/path/to/image3.png",
            predictions=None,
        ),
    ]
)

#
# Count the number of samples in the dataset
#

count = dataset.count()
print(count)  # the count

#
# Count the number of samples with `predictions`
#

count = dataset.count("predictions")
print(count)  # the count

#
# Count the number of objects in the `predictions` field
#

count = dataset.count("predictions.detections")
print(count)  # the count

#
# Count the number of objects in samples with > 2 predictions
#

count = dataset.count(
    (F("predictions.detections").length() > 2).if_else(
        F("predictions.detections"), None
    )
)
print(count)  # the count

Parameters

field_or_expr (None) –
a field name, embedded.field.name, fiftyone.core.expressions.ViewExpression, or MongoDB expression defining the field or expression to aggregate. If neither field_or_expr or expr is provided, the samples themselves are counted. This can also be a list or tuple of such arguments, in which case a tuple of corresponding aggregation results (each receiving the same additional keyword arguments, if any) will be returned
expr (None) –
a fiftyone.core.expressions.ViewExpression or MongoDB expression to apply to field_or_expr (which must be a field) before aggregating
safe (False) – whether to ignore nan/inf values when dealing with floating point values

Returns

the count

count_label_tags(label_fields=None)¶

Counts the occurrences of all label tags in the specified label field(s) of this collection.

Parameters: label_fields (None) – an optional name or iterable of names of fiftyone.core.labels.Label fields. By default, all label fields are used
Returns: a dict mapping tags to counts

count_sample_tags()¶

Counts the occurrences of sample tags in this collection.

Returns: a dict mapping tags to counts

count_values(field_or_expr, expr=None, safe=False)¶

Counts the occurrences of field values in the collection.

This aggregation is typically applied to countable field types (or lists of such types):

Examples:

import fiftyone as fo
from fiftyone import ViewField as F

dataset = fo.Dataset()
dataset.add_samples(
    [
        fo.Sample(
            filepath="/path/to/image1.png",
            tags=["sunny"],
            predictions=fo.Detections(
                detections=[
                    fo.Detection(label="cat"),
                    fo.Detection(label="dog"),
                ]
            ),
        ),
        fo.Sample(
            filepath="/path/to/image2.png",
            tags=["cloudy"],
            predictions=fo.Detections(
                detections=[
                    fo.Detection(label="cat"),
                    fo.Detection(label="rabbit"),
                ]
            ),
        ),
        fo.Sample(
            filepath="/path/to/image3.png",
            predictions=None,
        ),
    ]
)

#
# Compute the tag counts in the dataset
#

counts = dataset.count_values("tags")
print(counts)  # dict mapping values to counts

#
# Compute the predicted label counts in the dataset
#

counts = dataset.count_values("predictions.detections.label")
print(counts)  # dict mapping values to counts

#
# Compute the predicted label counts after some normalization
#

counts = dataset.count_values(
    F("predictions.detections.label").map_values(
        {"cat": "pet", "dog": "pet"}
    ).upper()
)
print(counts)  # dict mapping values to counts

Parameters

field_or_expr –
a field name, embedded.field.name, fiftyone.core.expressions.ViewExpression, or MongoDB expression defining the field or expression to aggregate. This can also be a list or tuple of such arguments, in which case a tuple of corresponding aggregation results (each receiving the same additional keyword arguments, if any) will be returned
expr (None) –
a fiftyone.core.expressions.ViewExpression or MongoDB expression to apply to field_or_expr (which must be a field) before aggregating
safe (False) – whether to treat nan/inf values as None when dealing with floating point values

Returns

a dict mapping values to counts

create_index(field_or_spec, unique=False, wait=True, **kwargs)¶

Creates an index on the given field or with the given specification, if necessary.

Indexes enable efficient sorting, merging, and other such operations.

Frame-level fields can be indexed by prepending "frames." to the field name.

Note

If an index with the same field(s) but different order(s) already exists, no new index will be created.

Use drop_index() to drop an existing index first if you wish to replace an existing index with new properties.

Note

If you are indexing a single field and it already has a unique constraint, it will be retained regardless of the unique value you specify. Conversely, if the given field already has a non-unique index but you requested a unique index, the existing index will be replaced with a unique index.

Use drop_index() to drop an existing index first if you wish to replace an existing index with new properties.

Parameters

field_or_spec – the field name, embedded.field.name, or index specification list. See pymongo.collection.Collection.create_index() for supported values
unique (False) – whether to add a uniqueness constraint to the index
wait (True) – whether to wait for index creation to finish
**kwargs – optional keyword arguments for pymongo.collection.Collection.create_index()

Returns

the name of the index

delete_annotation_run(anno_key)¶

Deletes the annotation run with the given key from this collection.

Calling this method only deletes the record of the annotation run from the collection; it will not delete any annotations loaded onto your dataset via load_annotations(), nor will it delete any associated information from the annotation backend.

Use load_annotation_results() to programmatically manage/delete a run from the annotation backend.

Parameters: anno_key – an annotation key

delete_annotation_runs()¶

Deletes all annotation runs from this collection.

Calling this method only deletes the records of the annotation runs from this collection; it will not delete any annotations loaded onto your dataset via load_annotations(), nor will it delete any associated information from the annotation backend.

Use load_annotation_results() to programmatically manage/delete runs in the annotation backend.

delete_brain_run(brain_key)¶

Deletes the brain method run with the given key from this collection.

Parameters: brain_key – a brain key

delete_brain_runs()¶: Deletes all brain method runs from this collection.

delete_evaluation(eval_key)¶

Deletes the evaluation results associated with the given evaluation key from this collection.

Parameters: eval_key – an evaluation key

delete_evaluations()¶: Deletes all evaluation results from this collection.

delete_run(run_key)¶

Deletes the run with the given key from this collection.

Parameters: run_key – a run key

delete_runs()¶: Deletes all runs from this collection.

distinct(field_or_expr, expr=None, safe=False)¶

Computes the distinct values of a field in the collection.

None-valued fields are ignored.

This aggregation is typically applied to countable field types (or lists of such types):

Examples:

import fiftyone as fo
from fiftyone import ViewField as F

dataset = fo.Dataset()
dataset.add_samples(
    [
        fo.Sample(
            filepath="/path/to/image1.png",
            tags=["sunny"],
            predictions=fo.Detections(
                detections=[
                    fo.Detection(label="cat"),
                    fo.Detection(label="dog"),
                ]
            ),
        ),
        fo.Sample(
            filepath="/path/to/image2.png",
            tags=["sunny", "cloudy"],
            predictions=fo.Detections(
                detections=[
                    fo.Detection(label="cat"),
                    fo.Detection(label="rabbit"),
                ]
            ),
        ),
        fo.Sample(
            filepath="/path/to/image3.png",
            predictions=None,
        ),
    ]
)

#
# Get the distinct tags in a dataset
#

values = dataset.distinct("tags")
print(values)  # list of distinct values

#
# Get the distinct predicted labels in a dataset
#

values = dataset.distinct("predictions.detections.label")
print(values)  # list of distinct values

#
# Get the distinct predicted labels after some normalization
#

values = dataset.distinct(
    F("predictions.detections.label").map_values(
        {"cat": "pet", "dog": "pet"}
    ).upper()
)
print(values)  # list of distinct values

Parameters

field_or_expr –
a field name, embedded.field.name, fiftyone.core.expressions.ViewExpression, or MongoDB expression defining the field or expression to aggregate. This can also be a list or tuple of such arguments, in which case a tuple of corresponding aggregation results (each receiving the same additional keyword arguments, if any) will be returned
expr (None) –
a fiftyone.core.expressions.ViewExpression or MongoDB expression to apply to field_or_expr (which must be a field) before aggregating
safe (False) – whether to ignore nan/inf values when dealing with floating point values

Returns

a sorted list of distinct values

draw_labels(output_dir, rel_dir=None, label_fields=None, overwrite=False, config=None, progress=None, **kwargs)¶

Renders annotated versions of the media in the collection with the specified label data overlaid to the given directory.

The filenames of the sample media are maintained, unless a name conflict would occur in output_dir, in which case an index of the form "-%d" % count is appended to the base filename.

Images are written in format fo.config.default_image_ext, and videos are written in format fo.config.default_video_ext.

Parameters

output_dir – the directory to write the annotated media
rel_dir (None) – an optional relative directory to strip from each input filepath to generate a unique identifier that is joined with output_dir to generate an output path for each annotated media. This argument allows for populating nested subdirectories in output_dir that match the shape of the input paths. The path is converted to an absolute path (if necessary) via fiftyone.core.storage.normalize_path()
label_fields (None) – a label field or list of label fields to render. By default, all fiftyone.core.labels.Label fields are drawn
overwrite (False) – whether to delete output_dir if it exists before rendering
config (None) – an optional fiftyone.utils.annotations.DrawConfig configuring how to draw the labels
progress (None) – whether to render a progress bar (True/False), use the default value fiftyone.config.show_progress_bars (None), or a progress callback function to invoke instead
**kwargs – optional keyword arguments specifying parameters of the default fiftyone.utils.annotations.DrawConfig to override

Returns

the list of paths to the rendered media

drop_index(field_or_name)¶

Drops the index for the given field or name, if necessary.

Parameters: field_or_name – a field name, embedded.field.name, or compound index name. Use list_indexes() to see the available indexes

evaluate_classifications(pred_field, gt_field='ground_truth', eval_key=None, classes=None, missing=None, method=None, progress=None, **kwargs)¶

Evaluates the classification predictions in this collection with respect to the specified ground truth labels.

By default, this method simply compares the ground truth and prediction for each sample, but other strategies such as binary evaluation and top-k matching can be configured via the method parameter.

You can customize the evaluation method by passing additional parameters for the method’s config class as kwargs.

The natively provided method values and their associated configs are:

"simple": fiftyone.utils.eval.classification.SimpleEvaluationConfig
"top-k": fiftyone.utils.eval.classification.TopKEvaluationConfig
"binary": fiftyone.utils.eval.classification.BinaryEvaluationConfig

If an eval_key is specified, then this method will record some statistics on each sample:

When evaluating sample-level fields, an eval_key field will be populated on each sample recording whether that sample’s prediction is correct.
When evaluating frame-level fields, an eval_key field will be populated on each frame recording whether that frame’s prediction is correct. In addition, an eval_key field will be populated on each sample that records the average accuracy of the frame predictions of the sample.

Parameters

pred_field – the name of the field containing the predicted fiftyone.core.labels.Classification instances
gt_field ("ground_truth") – the name of the field containing the ground truth fiftyone.core.labels.Classification instances
eval_key (None) – a string key to use to refer to this evaluation
classes (None) – the list of possible classes. If not provided, the observed ground truth/predicted labels are used
missing (None) – a missing label string. Any None-valued labels are given this label for results purposes
method (None) – a string specifying the evaluation method to use. The supported values are fo.evaluation_config.classification_backends.keys() and the default is fo.evaluation_config.classification_default_backend
progress (None) – whether to render a progress bar (True/False), use the default value fiftyone.config.show_progress_bars (None), or a progress callback function to invoke instead
**kwargs – optional keyword arguments for the constructor of the fiftyone.utils.eval.classification.ClassificationEvaluationConfig being used

Returns

a fiftyone.utils.eval.classification.ClassificationResults

evaluate_detections(pred_field, gt_field='ground_truth', eval_key=None, classes=None, missing=None, method=None, iou=0.5, use_masks=False, use_boxes=False, classwise=True, dynamic=True, progress=None, **kwargs)¶

Evaluates the specified predicted detections in this collection with respect to the specified ground truth detections.

This method supports evaluating the following spatial data types:

Object detections in fiftyone.core.labels.Detections format
Instance segmentations in fiftyone.core.labels.Detections format with their mask attributes populated
Polygons in fiftyone.core.labels.Polylines format
Keypoints in fiftyone.core.labels.Keypoints format
Temporal detections in fiftyone.core.labels.TemporalDetections format

For spatial object detection evaluation, this method uses COCO-style evaluation by default.

When evaluating keypoints, “IoUs” are computed via object keypoint similarity.

For temporal segment detection, this method uses ActivityNet-style evaluation by default.

You can use the method parameter to select a different method, and you can optionally customize the method by passing additional parameters for the method’s config class as kwargs.

The natively provided method values and their associated configs are:

"coco": fiftyone.utils.eval.coco.COCOEvaluationConfig
"open-images": fiftyone.utils.eval.openimages.OpenImagesEvaluationConfig
"activitynet": fiftyone.utils.eval.activitynet.ActivityNetEvaluationConfig

If an eval_key is provided, a number of fields are populated at the object- and sample-level recording the results of the evaluation:

True positive (TP), false positive (FP), and false negative (FN) counts for the each sample are saved in top-level fields of each sample:
```
TP: sample.<eval_key>_tp
FP: sample.<eval_key>_fp
FN: sample.<eval_key>_fn
```
In addition, when evaluating frame-level objects, TP/FP/FN counts are recorded for each frame:
```
TP: frame.<eval_key>_tp
FP: frame.<eval_key>_fp
FN: frame.<eval_key>_fn
```
The fields listed below are populated on each individual object; these fields tabulate the TP/FP/FN status of the object, the ID of the matching object (if any), and the matching IoU:
```
TP/FP/FN: object.<eval_key>
      ID: object.<eval_key>_id
     IoU: object.<eval_key>_iou
```

Parameters

pred_field – the name of the field containing the predicted fiftyone.core.labels.Detections, fiftyone.core.labels.Polylines, fiftyone.core.labels.Keypoints, or fiftyone.core.labels.TemporalDetections
gt_field ("ground_truth") – the name of the field containing the ground truth fiftyone.core.labels.Detections, fiftyone.core.labels.Polylines, fiftyone.core.labels.Keypoints, or fiftyone.core.labels.TemporalDetections
eval_key (None) – a string key to use to refer to this evaluation
classes (None) – the list of possible classes. If not provided, the observed ground truth/predicted labels are used
missing (None) – a missing label string. Any unmatched objects are given this label for results purposes
method (None) – a string specifying the evaluation method to use. The supported values are fo.evaluation_config.detection_backends.keys() and the default is fo.evaluation_config.detection_default_backend
iou (0.50) – the IoU threshold to use to determine matches
use_masks (False) – whether to compute IoUs using the instances masks in the mask attribute of the provided objects, which must be fiftyone.core.labels.Detection instances
use_boxes (False) – whether to compute IoUs using the bounding boxes of the provided fiftyone.core.labels.Polyline instances rather than using their actual geometries
classwise (True) – whether to only match objects with the same class label (True) or allow matches between classes (False)
dynamic (True) – whether to declare the dynamic object-level attributes that are populated on the dataset’s schema
progress (None) – whether to render a progress bar (True/False), use the default value fiftyone.config.show_progress_bars (None), or a progress callback function to invoke instead
**kwargs – optional keyword arguments for the constructor of the fiftyone.utils.eval.detection.DetectionEvaluationConfig being used

Returns

a fiftyone.utils.eval.detection.DetectionResults

evaluate_regressions(pred_field, gt_field='ground_truth', eval_key=None, missing=None, method=None, progress=None, **kwargs)¶

Evaluates the regression predictions in this collection with respect to the specified ground truth values.

You can customize the evaluation method by passing additional parameters for the method’s config class as kwargs.

The natively provided method values and their associated configs are:

"simple": fiftyone.utils.eval.regression.SimpleEvaluationConfig

If an eval_key is specified, then this method will record some statistics on each sample:

When evaluating sample-level fields, an eval_key field will be populated on each sample recording the error of that sample’s prediction.
When evaluating frame-level fields, an eval_key field will be populated on each frame recording the error of that frame’s prediction. In addition, an eval_key field will be populated on each sample that records the average error of the frame predictions of the sample.

Parameters

pred_field – the name of the field containing the predicted fiftyone.core.labels.Regression instances
gt_field ("ground_truth") – the name of the field containing the ground truth fiftyone.core.labels.Regression instances
eval_key (None) – a string key to use to refer to this evaluation
missing (None) – a missing value. Any None-valued regressions are given this value for results purposes
method (None) – a string specifying the evaluation method to use. The supported values are fo.evaluation_config.regression_backends.keys() and the default is fo.evaluation_config.regression_default_backend
progress (None) – whether to render a progress bar (True/False), use the default value fiftyone.config.show_progress_bars (None), or a progress callback function to invoke instead
**kwargs – optional keyword arguments for the constructor of the fiftyone.utils.eval.regression.RegressionEvaluationConfig being used

Returns

a fiftyone.utils.eval.regression.RegressionResults

evaluate_segmentations(pred_field, gt_field='ground_truth', eval_key=None, mask_targets=None, method=None, progress=None, **kwargs)¶

Evaluates the specified semantic segmentation masks in this collection with respect to the specified ground truth masks.

If the size of a predicted mask does not match the ground truth mask, it is resized to match the ground truth.

By default, this method simply performs pixelwise evaluation of the full masks, but other strategies such as boundary-only evaluation can be configured by passing additional parameters for the method’s config class as kwargs.

The natively provided method values and their associated configs are:

"simple": fiftyone.utils.eval.segmentation.SimpleEvaluationConfig

If an eval_key is provided, the accuracy, precision, and recall of each sample is recorded in top-level fields of each sample:

 Accuracy: sample.<eval_key>_accuracy
Precision: sample.<eval_key>_precision
   Recall: sample.<eval_key>_recall

In addition, when evaluating frame-level masks, the accuracy, precision, and recall of each frame if recorded in the following frame-level fields:

 Accuracy: frame.<eval_key>_accuracy
Precision: frame.<eval_key>_precision
   Recall: frame.<eval_key>_recall

Note

The mask values 0 and #000000 are treated as a background class for the purposes of computing evaluation metrics like precision and recall.

Parameters

pred_field – the name of the field containing the predicted fiftyone.core.labels.Segmentation instances
gt_field ("ground_truth") – the name of the field containing the ground truth fiftyone.core.labels.Segmentation instances
eval_key (None) – a string key to use to refer to this evaluation
mask_targets (None) – a dict mapping pixel values or RGB hex strings to labels. If not provided, the observed values are used as labels
method (None) – a string specifying the evaluation method to use. The supported values are fo.evaluation_config.segmentation_backends.keys() and the default is fo.evaluation_config.segmentation_default_backend
progress (None) – whether to render a progress bar (True/False), use the default value fiftyone.config.show_progress_bars (None), or a progress callback function to invoke instead
**kwargs – optional keyword arguments for the constructor of the fiftyone.utils.eval.segmentation.SegmentationEvaluationConfig being used

Returns

a fiftyone.utils.eval.segmentation.SegmentationResults

exclude(sample_ids)¶

Excludes the samples with the given IDs from the collection.

Examples:

import fiftyone as fo

dataset = fo.Dataset()
dataset.add_samples(
    [
        fo.Sample(filepath="/path/to/image1.png"),
        fo.Sample(filepath="/path/to/image2.png"),
        fo.Sample(filepath="/path/to/image3.png"),
    ]
)

#
# Exclude the first sample from the dataset
#

sample_id = dataset.first().id
view = dataset.exclude(sample_id)

#
# Exclude the first and last samples from the dataset
#

sample_ids = [dataset.first().id, dataset.last().id]
view = dataset.exclude(sample_ids)

Parameters

sample_ids –

the samples to exclude. Can be any of the following:

a sample ID
an iterable of sample IDs
a fiftyone.core.sample.Sample or fiftyone.core.sample.SampleView
an iterable of fiftyone.core.sample.Sample or fiftyone.core.sample.SampleView instances
a fiftyone.core.collections.SampleCollection

Returns

a fiftyone.core.view.DatasetView

exclude_by(field, values)¶

Excludes the samples with the given field values from the collection.

This stage is typically used to work with categorical fields (strings, ints, and bools). If you want to exclude samples based on floating point fields, use match().

Examples:

import fiftyone as fo

dataset = fo.Dataset()
dataset.add_samples(
    [
        fo.Sample(filepath="image%d.jpg" % i, int=i, str=str(i))
        for i in range(10)
    ]
)

#
# Create a view excluding samples whose `int` field have the given
# values
#

view = dataset.exclude_by("int", [1, 9, 3, 7, 5])
print(view.head(5))

#
# Create a view excluding samples whose `str` field have the given
# values
#

view = dataset.exclude_by("str", ["1", "9", "3", "7", "5"])
print(view.head(5))

Parameters

field – a field or embedded.field.name
values – a value or iterable of values to exclude by

Returns

a fiftyone.core.view.DatasetView

exclude_fields(field_names=None, meta_filter=None, _allow_missing=False)¶

Excludes the fields with the given names from the samples in the collection.

Note that default fields cannot be excluded.

Examples:

import fiftyone as fo

dataset = fo.Dataset()
dataset.add_samples(
    [
        fo.Sample(
            filepath="/path/to/image1.png",
            ground_truth=fo.Classification(label="cat"),
            predictions=fo.Classification(
                label="cat",
                confidence=0.9,
                mood="surly",
            ),
        ),
        fo.Sample(
            filepath="/path/to/image2.png",
            ground_truth=fo.Classification(label="dog"),
            predictions=fo.Classification(
                label="dog",
                confidence=0.8,
                mood="happy",
            ),
        ),
        fo.Sample(
            filepath="/path/to/image3.png",
        ),
    ]
)

#
# Exclude the `predictions` field from all samples
#

view = dataset.exclude_fields("predictions")

#
# Exclude the `mood` attribute from all classifications in the
# `predictions` field
#

view = dataset.exclude_fields("predictions.mood")

Parameters

field_names (None) – a field name or iterable of field names to exclude. May contain embedded.field.name as well
meta_filter (None) –
a filter that dynamically excludes fields in the collection’s schema according to the specified rule, which can be matched against the field’s name, type, description, and/or info. For example:
- Use meta_filter="2023" or meta_filter={"any": "2023"} to exclude fields that have the string “2023” anywhere in their name, type, description, or info
- Use meta_filter={"type": "StringField"} or meta_filter={"type": "Classification"} to exclude all string or classification fields, respectively
- Use meta_filter={"description": "my description"} to exclude fields whose description contains the string “my description”
- Use meta_filter={"info": "2023"} to exclude fields that have the string “2023” anywhere in their info
- Use meta_filter={"info.key": "value"}} to exclude fields that have a specific key/value pair in their info
- Include meta_filter={"include_nested_fields": True, ...} in your meta filter to include all nested fields in the filter

Returns

a fiftyone.core.view.DatasetView

exclude_frames(frame_ids, omit_empty=True)¶

Excludes the frames with the given IDs from the video collection.

Examples:

import fiftyone as fo
import fiftyone.zoo as foz

dataset = foz.load_zoo_dataset("quickstart-video")

#
# Exclude some specific frames
#

frame_ids = [
    dataset.first().frames.first().id,
    dataset.last().frames.last().id,
]

view = dataset.exclude_frames(frame_ids)

print(dataset.count("frames"))
print(view.count("frames"))

Parameters

frame_ids –
the frames to exclude. Can be any of the following:
- a frame ID
- an iterable of frame IDs
- a fiftyone.core.frame.Frame or fiftyone.core.frame.FrameView
- an iterable of fiftyone.core.frame.Frame or fiftyone.core.frame.FrameView instances
- a fiftyone.core.collections.SampleCollection whose frames to exclude
omit_empty (True) – whether to omit samples that have no frames after excluding the specified frames

Returns

a fiftyone.core.view.DatasetView

exclude_groups(group_ids)¶

Excludes the groups with the given IDs from the grouped collection.

Examples:

import fiftyone as fo
import fiftyone.zoo as foz

dataset = foz.load_zoo_dataset("quickstart-groups")

#
# Exclude some specific groups by ID
#

view = dataset.take(2)
group_ids = view.values("group.id")
other_groups = dataset.exclude_groups(group_ids)

assert len(set(group_ids) & set(other_groups.values("group.id"))) == 0

Parameters

groups_ids –

the groups to exclude. Can be any of the following:

a group ID
an iterable of group IDs
a fiftyone.core.sample.Sample or fiftyone.core.sample.SampleView
a group dict returned by get_group()
an iterable of fiftyone.core.sample.Sample or fiftyone.core.sample.SampleView instances
an iterable of group dicts returned by get_group()
a fiftyone.core.collections.SampleCollection

Returns

a fiftyone.core.view.DatasetView

exclude_labels(labels=None, ids=None, tags=None, fields=None, omit_empty=True)¶

Excludes the specified labels from the collection.

The returned view will omit samples, sample fields, and individual labels that do not match the specified selection criteria.

You can perform an exclusion via one or more of the following methods:

Provide the labels argument, which should contain a list of dicts in the format returned by fiftyone.core.session.Session.selected_labels, to exclude specific labels
Provide the ids argument to exclude labels with specific IDs
Provide the tags argument to exclude labels with specific tags

If multiple criteria are specified, labels must match all of them in order to be excluded.

By default, the exclusion is applied to all fiftyone.core.labels.Label fields, but you can provide the fields argument to explicitly define the field(s) in which to exclude.

Examples:

import fiftyone as fo
import fiftyone.zoo as foz

dataset = foz.load_zoo_dataset("quickstart")

#
# Exclude the labels currently selected in the App
#

session = fo.launch_app(dataset)

# Select some labels in the App...

view = dataset.exclude_labels(labels=session.selected_labels)

#
# Exclude labels with the specified IDs
#

# Grab some label IDs
ids = [
    dataset.first().ground_truth.detections[0].id,
    dataset.last().predictions.detections[0].id,
]

view = dataset.exclude_labels(ids=ids)

print(dataset.count("ground_truth.detections"))
print(view.count("ground_truth.detections"))

print(dataset.count("predictions.detections"))
print(view.count("predictions.detections"))

#
# Exclude labels with the specified tags
#

# Grab some label IDs
ids = [
    dataset.first().ground_truth.detections[0].id,
    dataset.last().predictions.detections[0].id,
]

# Give the labels a "test" tag
dataset = dataset.clone()  # create copy since we're modifying data
dataset.select_labels(ids=ids).tag_labels("test")

print(dataset.count_values("ground_truth.detections.tags"))
print(dataset.count_values("predictions.detections.tags"))

# Exclude the labels via their tag
view = dataset.exclude_labels(tags="test")

print(dataset.count("ground_truth.detections"))
print(view.count("ground_truth.detections"))

print(dataset.count("predictions.detections"))
print(view.count("predictions.detections"))

Parameters

labels (None) – a list of dicts specifying the labels to exclude in the format returned by fiftyone.core.session.Session.selected_labels
ids (None) – an ID or iterable of IDs of the labels to exclude
tags (None) – a tag or iterable of tags of labels to exclude
fields (None) – a field or iterable of fields from which to exclude
omit_empty (True) – whether to omit samples that have no labels after filtering

Returns

a fiftyone.core.view.DatasetView

exists(field, bool=None)¶

Returns a view containing the samples in the collection that have (or do not have) a non-None value for the given field or embedded field.

Examples:

import fiftyone as fo

dataset = fo.Dataset()
dataset.add_samples(
    [
        fo.Sample(
            filepath="/path/to/image1.png",
            ground_truth=fo.Classification(label="cat"),
            predictions=fo.Classification(label="cat", confidence=0.9),
        ),
        fo.Sample(
            filepath="/path/to/image2.png",
            ground_truth=fo.Classification(label="dog"),
            predictions=fo.Classification(label="dog", confidence=0.8),
        ),
        fo.Sample(
            filepath="/path/to/image3.png",
            ground_truth=fo.Classification(label="dog"),
            predictions=fo.Classification(label="dog"),
        ),
        fo.Sample(
            filepath="/path/to/image4.png",
            ground_truth=None,
            predictions=None,
        ),
        fo.Sample(filepath="/path/to/image5.png"),
    ]
)

#
# Only include samples that have a value in their `predictions`
# field
#

view = dataset.exists("predictions")

#
# Only include samples that do NOT have a value in their
# `predictions` field
#

view = dataset.exists("predictions", False)

#
# Only include samples that have prediction confidences
#

view = dataset.exists("predictions.confidence")

Parameters

field – the field name or embedded.field.name
bool (None) – whether to check if the field exists (None or True) or does not exist (False)

Returns

a fiftyone.core.view.DatasetView

export(export_dir=None, dataset_type=None, data_path=None, labels_path=None, export_media=None, rel_dir=None, dataset_exporter=None, label_field=None, frame_labels_field=None, overwrite=False, progress=None, **kwargs)¶

Exports the samples in the collection to disk.

You can perform exports with this method via the following basic patterns:

Provide export_dir and dataset_type to export the content to a directory in the default layout for the specified format, as documented in this page
Provide dataset_type along with data_path, labels_path, and/or export_media to directly specify where to export the source media and/or labels (if applicable) in your desired format. This syntax provides the flexibility to, for example, perform workflows like labels-only exports
Provide a dataset_exporter to which to feed samples to perform a fully-customized export

In all workflows, the remaining parameters of this method can be provided to further configure the export.

See this page for more information about the available export formats and examples of using this method.

See this guide for more details about exporting datasets in custom formats by defining your own fiftyone.utils.data.exporters.DatasetExporter.

This method will automatically coerce the data to match the requested export in the following cases:

When exporting in either an unlabeled image or image classification format, if a spatial label field is provided (fiftyone.core.labels.Detection, fiftyone.core.labels.Detections, fiftyone.core.labels.Polyline, or fiftyone.core.labels.Polylines), then the image patches of the provided samples will be exported
When exporting in labeled image dataset formats that expect list-type labels (fiftyone.core.labels.Classifications, fiftyone.core.labels.Detections, fiftyone.core.labels.Keypoints, or fiftyone.core.labels.Polylines), if a label field contains labels in non-list format (e.g., fiftyone.core.labels.Classification), the labels will be automatically upgraded to single-label lists
When exporting in labeled image dataset formats that expect fiftyone.core.labels.Detections labels, if a fiftyone.core.labels.Classification field is provided, the labels will be automatically upgraded to detections that span the entire images

Parameters

export_dir (None) –
the directory to which to export the samples in format dataset_type. This parameter may be omitted if you have provided appropriate values for the data_path and/or labels_path parameters. Alternatively, this can also be an archive path with one of the following extensions:
```
.zip, .tar, .tar.gz, .tgz, .tar.bz, .tbz
```
If an archive path is specified, the export is performed in a directory of same name (minus extension) and then automatically archived and the directory then deleted
dataset_type (None) – the fiftyone.types.Dataset type to write. If not specified, the default type for label_field is used
data_path (None) –
an optional parameter that enables explicit control over the location of the exported media for certain export formats. Can be any of the following:
- a folder name like "data" or "data/" specifying a subfolder of export_dir in which to export the media
- an absolute directory path in which to export the media. In this case, the export_dir has no effect on the location of the data
- a filename like "data.json" specifying the filename of a JSON manifest file in export_dir generated when export_media is "manifest"
- an absolute filepath specifying the location to write the JSON manifest file when export_media is "manifest". In this case, export_dir has no effect on the location of the data
If None, a default value of this parameter will be chosen based on the value of the export_media parameter. Note that this parameter is not applicable to certain export formats such as binary types like TF records
labels_path (None) –
an optional parameter that enables explicit control over the location of the exported labels. Only applicable when exporting in certain labeled dataset formats. Can be any of the following:
- a type-specific folder name like "labels" or "labels/" or a filename like "labels.json" or "labels.xml" specifying the location in export_dir in which to export the labels
- an absolute directory or filepath in which to export the labels. In this case, the export_dir has no effect on the location of the labels
For labeled datasets, the default value of this parameter will be chosen based on the export format so that the labels will be exported into export_dir
export_media (None) –
controls how to export the raw media. The supported values are:
- True: copy all media files into the output directory
- False: don’t export media. This option is only useful when exporting labeled datasets whose label format stores sufficient information to locate the associated media
- "move": move all media files into the output directory
- "symlink": create symlinks to the media files in the output directory
- "manifest": create a data.json in the output directory that maps UUIDs used in the labels files to the filepaths of the source media, rather than exporting the actual media
If None, an appropriate default value of this parameter will be chosen based on the value of the data_path parameter. Note that some dataset formats may not support certain values for this parameter (e.g., when exporting in binary formats such as TF records, “symlink” is not an option)
rel_dir (None) – an optional relative directory to strip from each input filepath to generate a unique identifier for each media. When exporting media, this identifier is joined with data_path to generate an output path for each exported media. This argument allows for populating nested subdirectories that match the shape of the input paths. The path is converted to an absolute path (if necessary) via fiftyone.core.storage.normalize_path()
dataset_exporter (None) – a fiftyone.utils.data.exporters.DatasetExporter to use to export the samples. When provided, parameters such as export_dir, dataset_type, data_path, and labels_path have no effect
label_field (None) –
controls the label field(s) to export. Only applicable to labeled datasets. Can be any of the following:
- the name of a label field to export
- a glob pattern of label field(s) to export
- a list or tuple of label field(s) to export
- a dictionary mapping label field names to keys to use when constructing the label dictionaries to pass to the exporter
Note that multiple fields can only be specified when the exporter used can handle dictionaries of labels. By default, the first field of compatible type for the exporter is used. When exporting labeled video datasets, this argument may contain frame fields prefixed by "frames."
frame_labels_field (None) –
controls the frame label field(s) to export. The "frames." prefix is optional. Only applicable to labeled video datasets. Can be any of the following:
- the name of a frame label field to export
- a glob pattern of frame label field(s) to export
- a list or tuple of frame label field(s) to export
- a dictionary mapping frame label field names to keys to use when constructing the frame label dictionaries to pass to the exporter
Note that multiple fields can only be specified when the exporter used can handle dictionaries of frame labels. By default, the first field of compatible type for the exporter is used
overwrite (False) – whether to delete existing directories before performing the export (True) or to merge the export with existing files and directories (False)
progress (None) – whether to render a progress bar (True/False), use the default value fiftyone.config.show_progress_bars (None), or a progress callback function to invoke instead
**kwargs – optional keyword arguments to pass to the dataset exporter’s constructor. If you are exporting image patches, this can also contain keyword arguments for fiftyone.utils.patches.ImagePatchesExtractor

filter_field(field, filter, only_matches=True)¶

Filters the values of a field or embedded field of each sample in the collection.

Values of field for which filter returns False are replaced with None.

Examples:

import fiftyone as fo
from fiftyone import ViewField as F

dataset = fo.Dataset()
dataset.add_samples(
    [
        fo.Sample(
            filepath="/path/to/image1.png",
            ground_truth=fo.Classification(label="cat"),
            predictions=fo.Classification(label="cat", confidence=0.9),
            numeric_field=1.0,
        ),
        fo.Sample(
            filepath="/path/to/image2.png",
            ground_truth=fo.Classification(label="dog"),
            predictions=fo.Classification(label="dog", confidence=0.8),
            numeric_field=-1.0,
        ),
        fo.Sample(
            filepath="/path/to/image3.png",
            ground_truth=None,
            predictions=None,
            numeric_field=None,
        ),
    ]
)

#
# Only include classifications in the `predictions` field
# whose `label` is "cat"
#

view = dataset.filter_field("predictions", F("label") == "cat")

#
# Only include samples whose `numeric_field` value is positive
#

view = dataset.filter_field("numeric_field", F() > 0)

Parameters

field – the field name or embedded.field.name
filter –
a fiftyone.core.expressions.ViewExpression or MongoDB expression that returns a boolean describing the filter to apply
only_matches (True) – whether to only include samples that match the filter (True) or include all samples (False)

Returns

a fiftyone.core.view.DatasetView

filter_keypoints(field, filter=None, labels=None, only_matches=True)¶

Filters the individual fiftyone.core.labels.Keypoint.points elements in the specified keypoints field of each sample in the collection.

Note

Use filter_labels() if you simply want to filter entire fiftyone.core.labels.Keypoint objects in a field.

Examples:

import fiftyone as fo
from fiftyone import ViewField as F

dataset = fo.Dataset()
dataset.add_samples(
    [
        fo.Sample(
            filepath="/path/to/image1.png",
            predictions=fo.Keypoints(
                keypoints=[
                    fo.Keypoint(
                        label="person",
                        points=[(0.1, 0.1), (0.1, 0.9), (0.9, 0.9), (0.9, 0.1)],
                        confidence=[0.7, 0.8, 0.95, 0.99],
                    )
                ]
            )
        ),
        fo.Sample(filepath="/path/to/image2.png"),
    ]
)

dataset.default_skeleton = fo.KeypointSkeleton(
    labels=["nose", "left eye", "right eye", "left ear", "right ear"],
    edges=[[0, 1, 2, 0], [0, 3], [0, 4]],
)

#
# Only include keypoints in the `predictions` field whose
# `confidence` is greater than 0.9
#

view = dataset.filter_keypoints(
    "predictions", filter=F("confidence") > 0.9
)

#
# Only include keypoints in the `predictions` field with less than
# four points
#

view = dataset.filter_keypoints(
    "predictions", labels=["left eye", "right eye"]
)

Parameters

field – the fiftyone.core.labels.Keypoint or fiftyone.core.labels.Keypoints field to filter
filter (None) –
a fiftyone.core.expressions.ViewExpression or MongoDB expression that returns a boolean, like F("confidence") > 0.5 or F("occluded") == False, to apply elementwise to the specified field, which must be a list of same length as fiftyone.core.labels.Keypoint.points
labels (None) – a label or iterable of keypoint skeleton labels to keep
only_matches (True) – whether to only include keypoints/samples with at least one point after filtering (True) or include all keypoints/samples (False)

Returns

a fiftyone.core.view.DatasetView

filter_labels(field, filter, only_matches=True, trajectories=False)¶

Filters the fiftyone.core.labels.Label field of each sample in the collection.

If the specified field is a single fiftyone.core.labels.Label type, fields for which filter returns False are replaced with None:

If the specified field is a fiftyone.core.labels.Label list type, the label elements for which filter returns False are omitted from the view:

Classifications Examples:

import fiftyone as fo
from fiftyone import ViewField as F

dataset = fo.Dataset()
dataset.add_samples(
    [
        fo.Sample(
            filepath="/path/to/image1.png",
            predictions=fo.Classification(label="cat", confidence=0.9),
        ),
        fo.Sample(
            filepath="/path/to/image2.png",
            predictions=fo.Classification(label="dog", confidence=0.8),
        ),
        fo.Sample(
            filepath="/path/to/image3.png",
            predictions=fo.Classification(label="rabbit"),
        ),
        fo.Sample(
            filepath="/path/to/image4.png",
            predictions=None,
        ),
    ]
)

#
# Only include classifications in the `predictions` field whose
# `confidence` is greater than 0.8
#

view = dataset.filter_labels("predictions", F("confidence") > 0.8)

#
# Only include classifications in the `predictions` field whose
# `label` is "cat" or "dog"
#

view = dataset.filter_labels(
    "predictions", F("label").is_in(["cat", "dog"])
)

Detections Examples:

import fiftyone as fo
from fiftyone import ViewField as F

dataset = fo.Dataset()
dataset.add_samples(
    [
        fo.Sample(
            filepath="/path/to/image1.png",
            predictions=fo.Detections(
                detections=[
                    fo.Detection(
                        label="cat",
                        bounding_box=[0.1, 0.1, 0.5, 0.5],
                        confidence=0.9,
                    ),
                    fo.Detection(
                        label="dog",
                        bounding_box=[0.2, 0.2, 0.3, 0.3],
                        confidence=0.8,
                    ),
                ]
            ),
        ),
        fo.Sample(
            filepath="/path/to/image2.png",
            predictions=fo.Detections(
                detections=[
                    fo.Detection(
                        label="cat",
                        bounding_box=[0.5, 0.5, 0.4, 0.4],
                        confidence=0.95,
                    ),
                    fo.Detection(label="rabbit"),
                ]
            ),
        ),
        fo.Sample(
            filepath="/path/to/image3.png",
            predictions=fo.Detections(
                detections=[
                    fo.Detection(
                        label="squirrel",
                        bounding_box=[0.25, 0.25, 0.5, 0.5],
                        confidence=0.5,
                    ),
                ]
            ),
        ),
        fo.Sample(
            filepath="/path/to/image4.png",
            predictions=None,
        ),
    ]
)

#
# Only include detections in the `predictions` field whose
# `confidence` is greater than 0.8
#

view = dataset.filter_labels("predictions", F("confidence") > 0.8)

#
# Only include detections in the `predictions` field whose `label`
# is "cat" or "dog"
#

view = dataset.filter_labels(
    "predictions", F("label").is_in(["cat", "dog"])
)

#
# Only include detections in the `predictions` field whose bounding
# box area is smaller than 0.2
#

# Bboxes are in [top-left-x, top-left-y, width, height] format
bbox_area = F("bounding_box")[2] * F("bounding_box")[3]

view = dataset.filter_labels("predictions", bbox_area < 0.2)

Polylines Examples:

import fiftyone as fo
from fiftyone import ViewField as F

dataset = fo.Dataset()
dataset.add_samples(
    [
        fo.Sample(
            filepath="/path/to/image1.png",
            predictions=fo.Polylines(
                polylines=[
                    fo.Polyline(
                        label="lane",
                        points=[[(0.1, 0.1), (0.1, 0.6)]],
                        filled=False,
                    ),
                    fo.Polyline(
                        label="road",
                        points=[[(0.2, 0.2), (0.5, 0.5), (0.2, 0.5)]],
                        filled=True,
                    ),
                ]
            ),
        ),
        fo.Sample(
            filepath="/path/to/image2.png",
            predictions=fo.Polylines(
                polylines=[
                    fo.Polyline(
                        label="lane",
                        points=[[(0.4, 0.4), (0.9, 0.4)]],
                        filled=False,
                    ),
                    fo.Polyline(
                        label="road",
                        points=[[(0.6, 0.6), (0.9, 0.9), (0.6, 0.9)]],
                        filled=True,
                    ),
                ]
            ),
        ),
        fo.Sample(
            filepath="/path/to/image3.png",
            predictions=None,
        ),
    ]
)

#
# Only include polylines in the `predictions` field that are filled
#

view = dataset.filter_labels("predictions", F("filled") == True)

#
# Only include polylines in the `predictions` field whose `label`
# is "lane"
#

view = dataset.filter_labels("predictions", F("label") == "lane")

#
# Only include polylines in the `predictions` field with at least
# 3 vertices
#

num_vertices = F("points").map(F().length()).sum()
view = dataset.filter_labels("predictions", num_vertices >= 3)

Keypoints Examples:

import fiftyone as fo
from fiftyone import ViewField as F

dataset = fo.Dataset()
dataset.add_samples(
    [
        fo.Sample(
            filepath="/path/to/image1.png",
            predictions=fo.Keypoint(
                label="house",
                points=[(0.1, 0.1), (0.1, 0.9), (0.9, 0.9), (0.9, 0.1)],
            ),
        ),
        fo.Sample(
            filepath="/path/to/image2.png",
            predictions=fo.Keypoint(
                label="window",
                points=[(0.4, 0.4), (0.5, 0.5), (0.6, 0.6)],
            ),
        ),
        fo.Sample(
            filepath="/path/to/image3.png",
            predictions=None,
        ),
    ]
)

#
# Only include keypoints in the `predictions` field whose `label`
# is "house"
#

view = dataset.filter_labels("predictions", F("label") == "house")

#
# Only include keypoints in the `predictions` field with less than
# four points
#

view = dataset.filter_labels("predictions", F("points").length() < 4)

Parameters

field – the label field to filter
filter –
a fiftyone.core.expressions.ViewExpression or MongoDB expression that returns a boolean describing the filter to apply
only_matches (True) – whether to only include samples with at least one label after filtering (True) or include all samples (False)
trajectories (False) – whether to match entire object trajectories for which the object matches the given filter on at least one frame. Only applicable to datasets that contain videos and frame-level label fields whose objects have their index attributes populated

Returns

a fiftyone.core.view.DatasetView

flatten(stages=None)¶

Returns a flattened view that contains all samples in the dynamic grouped collection.

Examples:

import fiftyone as fo
import fiftyone.zoo as foz
from fiftyone import ViewField as F

dataset = foz.load_zoo_dataset("cifar10", split="test")

# Group samples by ground truth label
grouped_view = dataset.take(1000).group_by("ground_truth.label")
print(len(grouped_view))  # 10

# Return a flat view that contains 10 samples from each class
flat_view = grouped_view.flatten(fo.Limit(10))
print(len(flat_view))  # 100

Parameters: stages (None) – a fiftyone.core.stages.ViewStage or list of fiftyone.core.stages.ViewStage instances to apply to each group’s samples while flattening
Returns: a fiftyone.core.view.DatasetView

geo_near(point, location_field=None, min_distance=None, max_distance=None, query=None)¶

Sorts the samples in the collection by their proximity to a specified geolocation.

Note

This stage must be the first stage in any fiftyone.core.view.DatasetView in which it appears.

Examples:

import fiftyone as fo
import fiftyone.zoo as foz

TIMES_SQUARE = [-73.9855, 40.7580]

dataset = foz.load_zoo_dataset("quickstart-geo")

#
# Sort the samples by their proximity to Times Square
#

view = dataset.geo_near(TIMES_SQUARE)

#
# Sort the samples by their proximity to Times Square, and only
# include samples within 5km
#

view = dataset.geo_near(TIMES_SQUARE, max_distance=5000)

#
# Sort the samples by their proximity to Times Square, and only
# include samples that are in Manhattan
#

import fiftyone.utils.geojson as foug

in_manhattan = foug.geo_within(
    "location.point",
    [
        [
            [-73.949701, 40.834487],
            [-73.896611, 40.815076],
            [-73.998083, 40.696534],
            [-74.031751, 40.715273],
            [-73.949701, 40.834487],
        ]
    ]
)

view = dataset.geo_near(
    TIMES_SQUARE, location_field="location", query=in_manhattan
)

Parameters

point –
the reference point to compute distances to. Can be any of the following:
- A [longitude, latitude] list
- A GeoJSON dict with Point type
- A fiftyone.core.labels.GeoLocation instance whose point attribute contains the point
location_field (None) –
the location data of each sample to use. Can be any of the following:
- The name of a fiftyone.core.fields.GeoLocation field whose point attribute to use as location data
- An embedded.field.name containing GeoJSON data to use as location data
- None, in which case there must be a single fiftyone.core.fields.GeoLocation field on the samples, which is used by default
min_distance (None) – filter samples that are less than this distance (in meters) from point
max_distance (None) – filter samples that are greater than this distance (in meters) from point
query (None) – an optional dict defining a MongoDB read query that samples must match in order to be included in this view

Returns

a fiftyone.core.view.DatasetView

geo_within(boundary, location_field=None, strict=True)¶

Filters the samples in this collection to only include samples whose geolocation is within a specified boundary.

Examples:

import fiftyone as fo
import fiftyone.zoo as foz

MANHATTAN = [
    [
        [-73.949701, 40.834487],
        [-73.896611, 40.815076],
        [-73.998083, 40.696534],
        [-74.031751, 40.715273],
        [-73.949701, 40.834487],
    ]
]

dataset = foz.load_zoo_dataset("quickstart-geo")

#
# Create a view that only contains samples in Manhattan
#

view = dataset.geo_within(MANHATTAN)

Parameters

boundary – a fiftyone.core.labels.GeoLocation, fiftyone.core.labels.GeoLocations, GeoJSON dict, or list of coordinates that define a Polygon or MultiPolygon to search within
location_field (None) –
the location data of each sample to use. Can be any of the following:
- The name of a fiftyone.core.fields.GeoLocation field whose point attribute to use as location data
- An embedded.field.name that directly contains the GeoJSON location data to use
- None, in which case there must be a single fiftyone.core.fields.GeoLocation field on the samples, which is used by default
strict (True) – whether a sample’s location data must strictly fall within boundary (True) in order to match, or whether any intersection suffices (False)

Returns

a fiftyone.core.view.DatasetView

get_annotation_info(anno_key)¶

Returns information about the annotation run with the given key on this collection.

Parameters: anno_key – an annotation key
Returns: a fiftyone.core.annotation.AnnotationInfo

get_brain_info(brain_key)¶

Returns information about the brain method run with the given key on this collection.

Parameters: brain_key – a brain key
Returns: a fiftyone.core.brain.BrainInfo

get_classes(field)¶

Gets the classes list for the given field, or None if no classes are available.

Classes are first retrieved from classes() if they exist, otherwise from default_classes().

Parameters: field – a field name
Returns: a list of classes, or None

get_dynamic_field_schema(fields=None, recursive=True)¶

Returns a schema dictionary describing the dynamic fields of the samples in the collection.

Dynamic fields are embedded document fields with at least one non-None value that have not been declared on the dataset’s schema.

Parameters

fields (None) – an optional field or iterable of fields for which to return dynamic fields. By default, all fields are considered
recursive (True) – whether to recursively inspect nested lists and embedded documents

Returns

a dict mapping field paths to fiftyone.core.fields.Field instances or lists of them

get_dynamic_frame_field_schema(fields=None, recursive=True)¶

Returns a schema dictionary describing the dynamic fields of the frames in the collection.

Dynamic fields are embedded document fields with at least one non-None value that have not been declared on the dataset’s schema.

Parameters

fields (None) – an optional field or iterable of fields for which to return dynamic fields. By default, all fields are considered
recursive (True) – whether to recursively inspect nested lists and embedded documents

Returns

a dict mapping field paths to fiftyone.core.fields.Field instances or lists of them, or None if the collection does not contain videos

get_evaluation_info(eval_key)¶

Returns information about the evaluation with the given key on this collection.

Parameters: eval_key – an evaluation key
Returns: an fiftyone.core.evaluation.EvaluationInfo

get_field(path, ftype=None, embedded_doc_type=None, read_only=None, include_private=False, leaf=False)¶

Returns the field instance of the provided path, or None if one does not exist.

Parameters

path – a field path
ftype (None) – an optional field type to enforce. Must be a subclass of fiftyone.core.fields.Field
embedded_doc_type (None) – an optional embedded document type to enforce. Must be a subclass of fiftyone.core.odm.BaseEmbeddedDocument
read_only (None) – whether to optionally enforce that the field is read-only (True) or not read-only (False)
include_private (False) – whether to include fields that start with _ in the returned schema
leaf (False) – whether to return the subfield of list fields

Returns

a fiftyone.core.fields.Field instance or None

Raises

ValueError – if the field does not match provided constraints

get_index_information(include_stats=False)¶

Returns a dictionary of information about the indexes on this collection.

See pymongo.collection.Collection.index_information() for details on the structure of this dictionary.

Parameters: include_stats (False) – whether to include the size and build status of each index
Returns: a dict mapping index names to info dicts

get_mask_targets(field)¶

Gets the mask targets for the given field, or None if no mask targets are available.

Mask targets are first retrieved from mask_targets() if they exist, otherwise from default_mask_targets().

Parameters: field – a field name
Returns: a list of classes, or None

get_run_info(run_key)¶

Returns information about the run with the given key on this collection.

Parameters: run_key – a run key
Returns: a fiftyone.core.runs.RunInfo

get_skeleton(field)¶

Gets the keypoint skeleton for the given field, or None if no skeleton is available.

Skeletons are first retrieved from skeletons() if they exist, otherwise from default_skeleton().

Parameters: field – a field name
Returns: a list of classes, or None

group_by(field_or_expr, order_by=None, reverse=False, flat=False, match_expr=None, sort_expr=None, create_index=True)¶

Creates a view that groups the samples in the collection by a specified field or expression.

Examples:

import fiftyone as fo
import fiftyone.zoo as foz
from fiftyone import ViewField as F

dataset = foz.load_zoo_dataset("cifar10", split="test")

#
# Take 1000 samples at random and group them by ground truth label
#

view = dataset.take(1000).group_by("ground_truth.label")

for group in view.iter_dynamic_groups():
    group_value = group.first().ground_truth.label
    print("%s: %d" % (group_value, len(group)))

#
# Variation of above operation that arranges the groups in
# decreasing order of size and immediately flattens them
#

from itertools import groupby

view = dataset.take(1000).group_by(
    "ground_truth.label",
    flat=True,
    sort_expr=F().length(),
    reverse=True,
)

rle = lambda v: [(k, len(list(g))) for k, g in groupby(v)]
for label, count in rle(view.values("ground_truth.label")):
    print("%s: %d" % (label, count))

Parameters

field_or_expr – the field or embedded.field.name to group by, or a list of field names defining a compound group key, or a fiftyone.core.expressions.ViewExpression or MongoDB aggregation expression that defines the value to group by
order_by (None) – an optional field by which to order the samples in each group
reverse (False) – whether to return the results in descending order. Applies both to order_by and sort_expr
flat (False) – whether to return a grouped collection (False) or a flattened collection (True)
match_expr (None) –
an optional fiftyone.core.expressions.ViewExpression or MongoDB aggregation expression that defines which groups to include in the output view. If provided, this expression will be evaluated on the list of samples in each group. Only applicable when flat=True
sort_expr (None) –
an optional fiftyone.core.expressions.ViewExpression or MongoDB aggregation expression that defines how to sort the groups in the output view. If provided, this expression will be evaluated on the list of samples in each group. Only applicable when flat=True
create_index (True) – whether to create an index, if necessary, to optimize the grouping. Only applicable when grouping by field(s), not expressions

Returns

a fiftyone.core.view.DatasetView

has_annotation_run(anno_key)¶

Whether this collection has an annotation run with the given key.

Parameters: anno_key – an annotation key
Returns: True/False

property has_annotation_runs¶: Whether this collection has any annotation runs.

has_brain_run(brain_key)¶

Whether this collection has a brain method run with the given key.

Parameters: brain_key – a brain key
Returns: True/False

property has_brain_runs¶: Whether this collection has any brain runs.

has_classes(field)¶

Determines whether this collection has a classes list for the given field.

Classes may be defined either in classes() or default_classes().

Parameters: field – a field name
Returns: True/False

has_evaluation(eval_key)¶

Whether this collection has an evaluation with the given key.

Parameters: eval_key – an evaluation key
Returns: True/False

property has_evaluations¶: Whether this collection has any evaluation results.

has_field(path)¶

Determines whether the collection has a field with the given name.

Parameters: path – the field name or embedded.field.name
Returns: True/False

has_frame_field(path)¶

Determines whether the collection has a frame-level field with the given name.

Parameters: path – the field name or embedded.field.name
Returns: True/False

has_mask_targets(field)¶

Determines whether this collection has mask targets for the given field.

Mask targets may be defined either in mask_targets() or default_mask_targets().

Parameters: field – a field name
Returns: True/False

has_run(run_key)¶

Whether this collection has a run with the given key.

Parameters: run_key – a run key
Returns: True/False

property has_runs¶: Whether this collection has any runs.

has_sample_field(path)¶

Determines whether the collection has a sample field with the given name.

Parameters: path – the field name or embedded.field.name
Returns: True/False

has_skeleton(field)¶

Determines whether this collection has a keypoint skeleton for the given field.

Keypoint skeletons may be defined either in skeletons() or default_skeleton().

Parameters: field – a field name
Returns: True/False

histogram_values(field_or_expr, expr=None, bins=None, range=None, auto=False)¶

Computes a histogram of the field values in the collection.

This aggregation is typically applied to numeric field types (or lists of such types):

Examples:

import numpy as np
import matplotlib.pyplot as plt

import fiftyone as fo
from fiftyone import ViewField as F

samples = []
for idx in range(100):
    samples.append(
        fo.Sample(
            filepath="/path/to/image%d.png" % idx,
            numeric_field=np.random.randn(),
            numeric_list_field=list(np.random.randn(10)),
        )
    )

dataset = fo.Dataset()
dataset.add_samples(samples)

def plot_hist(counts, edges):
    counts = np.asarray(counts)
    edges = np.asarray(edges)
    left_edges = edges[:-1]
    widths = edges[1:] - edges[:-1]
    plt.bar(left_edges, counts, width=widths, align="edge")

#
# Compute a histogram of a numeric field
#

counts, edges, other = dataset.histogram_values(
    "numeric_field", bins=50, range=(-4, 4)
)

plot_hist(counts, edges)
plt.show(block=False)

#
# Compute the histogram of a numeric list field
#

counts, edges, other = dataset.histogram_values(
    "numeric_list_field", bins=50
)

plot_hist(counts, edges)
plt.show(block=False)

#
# Compute the histogram of a transformation of a numeric field
#

counts, edges, other = dataset.histogram_values(
    2 * (F("numeric_field") + 1), bins=50
)

plot_hist(counts, edges)
plt.show(block=False)

Parameters

field_or_expr –
a field name, embedded.field.name, fiftyone.core.expressions.ViewExpression, or MongoDB expression defining the field or expression to aggregate. This can also be a list or tuple of such arguments, in which case a tuple of corresponding aggregation results (each receiving the same additional keyword arguments, if any) will be returned
expr (None) –
a fiftyone.core.expressions.ViewExpression or MongoDB expression to apply to field_or_expr (which must be a field) before aggregating
bins (None) – can be either an integer number of bins to generate or a monotonically increasing sequence specifying the bin edges to use. By default, 10 bins are created. If bins is an integer and no range is specified, bin edges are automatically distributed in an attempt to evenly distribute the counts in each bin
range (None) – a (lower, upper) tuple specifying a range in which to generate equal-width bins. Only applicable when bins is an integer
auto (False) – whether to automatically choose bin edges in an attempt to evenly distribute the counts in each bin. If this option is chosen, bins will only be used if it is an integer, and the range parameter is ignored

Returns

a tuple of

counts: a list of counts in each bin
edges: an increasing list of bin edges of length len(counts) + 1. Note that each bin is treated as having an inclusive lower boundary and exclusive upper boundary, [lower, upper), including the rightmost bin
other: the number of items outside the bins

init_run(**kwargs)¶

Initializes a config instance for a new run.

Parameters: **kwargs – JSON serializable config parameters
Returns: a fiftyone.core.runs.RunConfig

init_run_results(run_key, **kwargs)¶

Initializes a results instance for the run with the given key.

Parameters

run_key – a run key
**kwargs – JSON serializable data

Returns

a fiftyone.core.runs.RunResults

limit(limit)¶

Returns a view with at most the given number of samples.

Examples:

import fiftyone as fo

dataset = fo.Dataset()
dataset.add_samples(
    [
        fo.Sample(
            filepath="/path/to/image1.png",
            ground_truth=fo.Classification(label="cat"),
        ),
        fo.Sample(
            filepath="/path/to/image2.png",
            ground_truth=fo.Classification(label="dog"),
        ),
        fo.Sample(
            filepath="/path/to/image3.png",
            ground_truth=None,
        ),
    ]
)

#
# Only include the first 2 samples in the view
#

view = dataset.limit(2)

Parameters: limit – the maximum number of samples to return. If a non-positive number is provided, an empty view is returned
Returns: a fiftyone.core.view.DatasetView

limit_labels(field, limit)¶

Limits the number of fiftyone.core.labels.Label instances in the specified labels list field of each sample in the collection.

The specified field must be one of the following types:

Examples:

import fiftyone as fo
from fiftyone import ViewField as F

dataset = fo.Dataset()
dataset.add_samples(
    [
        fo.Sample(
            filepath="/path/to/image1.png",
            predictions=fo.Detections(
                detections=[
                    fo.Detection(
                        label="cat",
                        bounding_box=[0.1, 0.1, 0.5, 0.5],
                        confidence=0.9,
                    ),
                    fo.Detection(
                        label="dog",
                        bounding_box=[0.2, 0.2, 0.3, 0.3],
                        confidence=0.8,
                    ),
                ]
            ),
        ),
        fo.Sample(
            filepath="/path/to/image2.png",
            predictions=fo.Detections(
                detections=[
                    fo.Detection(
                        label="cat",
                        bounding_box=[0.5, 0.5, 0.4, 0.4],
                        confidence=0.95,
                    ),
                    fo.Detection(label="rabbit"),
                ]
            ),
        ),
        fo.Sample(
            filepath="/path/to/image4.png",
            predictions=None,
        ),
    ]
)

#
# Only include the first detection in the `predictions` field of
# each sample
#

view = dataset.limit_labels("predictions", 1)

Parameters

field – the labels list field to filter
limit – the maximum number of labels to include in each labels list. If a non-positive number is provided, all lists will be empty

Returns

a fiftyone.core.view.DatasetView

classmethod list_aggregations()¶

Returns a list of all available methods on this collection that apply fiftyone.core.aggregations.Aggregation operations to this collection.

Returns: a list of SampleCollection method names

list_annotation_runs(type=None, method=None, **kwargs)¶

Returns a list of annotation keys on this collection.

Parameters

type (None) –
a specific annotation run type to match, which can be:
- a string fiftyone.core.annotations.AnnotationMethodConfig.type
- a fiftyone.core.annotations.AnnotationMethod class or its fully-qualified class name string
method (None) – a specific fiftyone.core.annotations.AnnotationMethodConfig.method string to match
**kwargs – optional config parameters to match

Returns

a list of annotation keys

list_brain_runs(type=None, method=None, **kwargs)¶

Returns a list of brain keys on this collection.

Parameters

type (None) –
a specific brain run type to match, which can be:
- a string fiftyone.core.brain.BrainMethodConfig.type
- a fiftyone.core.brain.BrainMethod class or its fully-qualified class name string
method (None) – a specific fiftyone.core.brain.BrainMethodConfig.method string to match
**kwargs – optional config parameters to match

Returns

a list of brain keys

list_evaluations(type=None, method=None, **kwargs)¶

Returns a list of evaluation keys on this collection.

Parameters

type (None) –
a specific evaluation type to match, which can be:
- a string fiftyone.core.evaluations.EvaluationMethodConfig.type
- a fiftyone.core.evaluations.EvaluationMethod class or its fully-qualified class name string
method (None) – a specific fiftyone.core.evaluations.EvaluationMethodConfig.method string to match
**kwargs – optional config parameters to match

Returns

a list of evaluation keys

list_indexes()¶

Returns the list of index names on this collection.

Single-field indexes are referenced by their field name, while compound indexes are referenced by more complicated strings. See pymongo.collection.Collection.index_information() for details on the compound format.

Returns: the list of index names

list_runs(**kwargs)¶

Returns a list of run keys on this collection.

Parameters: **kwargs – optional config parameters to match
Returns: a list of run keys

list_schema(field_or_expr, expr=None)¶

Extracts the value type(s) in a specified list field across all samples in the collection.

Examples:

from datetime import datetime
import fiftyone as fo

dataset = fo.Dataset()

sample1 = fo.Sample(
    filepath="image1.png",
    ground_truth=fo.Classification(
        label="cat",
        info=[
            fo.DynamicEmbeddedDocument(
                task="initial_annotation",
                author="Alice",
                timestamp=datetime(1970, 1, 1),
                notes=["foo", "bar"],
            ),
            fo.DynamicEmbeddedDocument(
                task="editing_pass",
                author="Bob",
                timestamp=datetime.utcnow(),
            ),
        ],
    ),
)

sample2 = fo.Sample(
    filepath="image2.png",
    ground_truth=fo.Classification(
        label="dog",
        info=[
            fo.DynamicEmbeddedDocument(
                task="initial_annotation",
                author="Bob",
                timestamp=datetime(2018, 10, 18),
                notes=["spam", "eggs"],
            ),
        ],
    ),
)

dataset.add_samples([sample1, sample2])

# Determine that `ground_truth.info` contains embedded documents
print(dataset.list_schema("ground_truth.info"))
# fo.EmbeddedDocumentField

# Determine the fields of the embedded documents in the list
print(dataset.schema("ground_truth.info[]"))
# {'task': StringField, ..., 'notes': ListField}

# Determine the type of the values in the nested `notes` list field
# Since `ground_truth.info` is not yet declared on the dataset's
# schema, we must manually include `[]` to unwind the info lists
print(dataset.list_schema("ground_truth.info[].notes"))
# fo.StringField

# Declare the `ground_truth.info` field
dataset.add_sample_field(
    "ground_truth.info",
    fo.ListField,
    subfield=fo.EmbeddedDocumentField,
    embedded_doc_type=fo.DynamicEmbeddedDocument,
)

# Now we can inspect the nested `notes` field without unwinding
print(dataset.list_schema("ground_truth.info.notes"))
# fo.StringField

Parameters

field_or_expr –
a field name, embedded.field.name, fiftyone.core.expressions.ViewExpression, or MongoDB expression defining the field or expression to aggregate
expr (None) –
a fiftyone.core.expressions.ViewExpression or MongoDB expression to apply to field_or_expr (which must be a field) before aggregating

Returns

a fiftyone.core.fields.Field or list of fiftyone.core.fields.Field instances describing the value type(s) in the list

classmethod list_view_stages()¶

Returns a list of all available methods on this collection that apply fiftyone.core.stages.ViewStage operations to this collection.

Returns: a list of SampleCollection method names

load_annotation_results(anno_key, cache=True, **kwargs)¶

Loads the results for the annotation run with the given key on this collection.

The fiftyone.utils.annotations.AnnotationResults object returned by this method will provide a variety of backend-specific methods allowing you to perform actions such as checking the status and deleting this run from the annotation backend.

Use load_annotations() to load the labels from an annotation run onto your FiftyOne dataset.

Parameters

anno_key – an annotation key
cache (True) – whether to cache the results on the collection
**kwargs – keyword arguments for run’s fiftyone.core.annotation.AnnotationMethodConfig.load_credentials() method

Returns

a fiftyone.utils.annotations.AnnotationResults

load_annotation_view(anno_key, select_fields=False)¶

Loads the fiftyone.core.view.DatasetView on which the specified annotation run was performed on this collection.

Parameters

anno_key – an annotation key
select_fields (False) – whether to exclude fields involved in other annotation runs

Returns

a fiftyone.core.view.DatasetView

load_annotations(anno_key, dest_field=None, unexpected='prompt', cleanup=False, progress=None, **kwargs)¶

Downloads the labels from the given annotation run from the annotation backend and merges them into this collection.

See this page for more information about using this method to import annotations that you have scheduled by calling annotate().

Parameters

anno_key – an annotation key
dest_field (None) – an optional name of a new destination field into which to load the annotations, or a dict mapping field names in the run’s label schema to new destination field names
unexpected ("prompt") –
how to deal with any unexpected labels that don’t match the run’s label schema when importing. The supported values are:
- "prompt": present an interactive prompt to direct/discard unexpected labels
- "ignore": automatically ignore any unexpected labels
- "keep": automatically keep all unexpected labels in a field whose name matches the the label type
- "return": return a dict containing all unexpected labels, or None if there aren’t any
cleanup (False) – whether to delete any informtation regarding this run from the annotation backend after loading the annotations
progress (None) – whether to render a progress bar (True/False), use the default value fiftyone.config.show_progress_bars (None), or a progress callback function to invoke instead
**kwargs – keyword arguments for the run’s fiftyone.core.annotation.AnnotationMethodConfig.load_credentials() method

Returns

None, unless unexpected=="return" and unexpected labels are found, in which case a dict containing the extra labels is returned

load_brain_results(brain_key, cache=True, load_view=True, **kwargs)¶

Loads the results for the brain method run with the given key on this collection.

Parameters

brain_key – a brain key
cache (True) – whether to cache the results on the collection
load_view (True) – whether to load the view on which the results were computed (True) or the full dataset (False)
**kwargs – keyword arguments for the run’s fiftyone.core.brain.BrainMethodConfig.load_credentials() method

Returns

a fiftyone.core.brain.BrainResults

load_brain_view(brain_key, select_fields=False)¶

Loads the fiftyone.core.view.DatasetView on which the specified brain method run was performed on this collection.

Parameters

brain_key – a brain key
select_fields (False) – whether to exclude fields involved in other brain method runs

Returns

a fiftyone.core.view.DatasetView

load_evaluation_results(eval_key, cache=True, **kwargs)¶

Loads the results for the evaluation with the given key on this collection.

Parameters

eval_key – an evaluation key
cache (True) – whether to cache the results on the collection
**kwargs – keyword arguments for the run’s fiftyone.core.evaluation.EvaluationMethodConfig.load_credentials() method

Returns

a fiftyone.core.evaluation.EvaluationResults

load_evaluation_view(eval_key, select_fields=False)¶

Loads the fiftyone.core.view.DatasetView on which the specified evaluation was performed on this collection.

Parameters

eval_key – an evaluation key
select_fields (False) – whether to exclude fields involved in other evaluations

Returns

a fiftyone.core.view.DatasetView

load_run_results(run_key, cache=True, load_view=True, **kwargs)¶

Loads the results for the run with the given key on this collection.

Parameters

run_key – a run key
cache (True) – whether to cache the results on the collection
load_view (True) – whether to load the view on which the results were computed (True) or the full dataset (False)
**kwargs – keyword arguments for the run’s fiftyone.core.runs.RunConfig.load_credentials() method

Returns

a fiftyone.core.runs.RunResults

load_run_view(run_key, select_fields=False)¶

Loads the fiftyone.core.view.DatasetView on which the specified run was performed on this collection.

Parameters

run_key – a run key
select_fields (False) – whether to exclude fields involved in other runs

Returns

a fiftyone.core.view.DatasetView

make_unique_field_name(root='')¶

Makes a unique field name with the given root name for the collection.

Parameters: root – an optional root for the output field name
Returns: the field name

map_labels(field, map)¶

Maps the label values of a fiftyone.core.labels.Label field to new values for each sample in the collection.

Examples:

import fiftyone as fo
from fiftyone import ViewField as F

dataset = fo.Dataset()
dataset.add_samples(
    [
        fo.Sample(
            filepath="/path/to/image1.png",
            weather=fo.Classification(label="sunny"),
            predictions=fo.Detections(
                detections=[
                    fo.Detection(
                        label="cat",
                        bounding_box=[0.1, 0.1, 0.5, 0.5],
                        confidence=0.9,
                    ),
                    fo.Detection(
                        label="dog",
                        bounding_box=[0.2, 0.2, 0.3, 0.3],
                        confidence=0.8,
                    ),
                ]
            ),
        ),
        fo.Sample(
            filepath="/path/to/image2.png",
            weather=fo.Classification(label="cloudy"),
            predictions=fo.Detections(
                detections=[
                    fo.Detection(
                        label="cat",
                        bounding_box=[0.5, 0.5, 0.4, 0.4],
                        confidence=0.95,
                    ),
                    fo.Detection(label="rabbit"),
                ]
            ),
        ),
        fo.Sample(
            filepath="/path/to/image3.png",
            weather=fo.Classification(label="partly cloudy"),
            predictions=fo.Detections(
                detections=[
                    fo.Detection(
                        label="squirrel",
                        bounding_box=[0.25, 0.25, 0.5, 0.5],
                        confidence=0.5,
                    ),
                ]
            ),
        ),
        fo.Sample(
            filepath="/path/to/image4.png",
            predictions=None,
        ),
    ]
)

#
# Map the "partly cloudy" weather label to "cloudy"
#

view = dataset.map_labels("weather", {"partly cloudy": "cloudy"})

#
# Map "rabbit" and "squirrel" predictions to "other"
#

view = dataset.map_labels(
    "predictions", {"rabbit": "other", "squirrel": "other"}
)

Parameters

field – the labels field to map
map – a dict mapping label values to new label values

Returns

a fiftyone.core.view.DatasetView

match(filter)¶

Filters the samples in the collection by the given filter.

Examples:

import fiftyone as fo
from fiftyone import ViewField as F

dataset = fo.Dataset()
dataset.add_samples(
    [
        fo.Sample(
            filepath="/path/to/image1.png",
            weather=fo.Classification(label="sunny"),
            predictions=fo.Detections(
                detections=[
                    fo.Detection(
                        label="cat",
                        bounding_box=[0.1, 0.1, 0.5, 0.5],
                        confidence=0.9,
                    ),
                    fo.Detection(
                        label="dog",
                        bounding_box=[0.2, 0.2, 0.3, 0.3],
                        confidence=0.8,
                    ),
                ]
            ),
        ),
        fo.Sample(
            filepath="/path/to/image2.jpg",
            weather=fo.Classification(label="cloudy"),
            predictions=fo.Detections(
                detections=[
                    fo.Detection(
                        label="cat",
                        bounding_box=[0.5, 0.5, 0.4, 0.4],
                        confidence=0.95,
                    ),
                    fo.Detection(label="rabbit"),
                ]
            ),
        ),
        fo.Sample(
            filepath="/path/to/image3.png",
            weather=fo.Classification(label="partly cloudy"),
            predictions=fo.Detections(
                detections=[
                    fo.Detection(
                        label="squirrel",
                        bounding_box=[0.25, 0.25, 0.5, 0.5],
                        confidence=0.5,
                    ),
                ]
            ),
        ),
        fo.Sample(
            filepath="/path/to/image4.jpg",
            predictions=None,
        ),
    ]
)

#
# Only include samples whose `filepath` ends with ".jpg"
#

view = dataset.match(F("filepath").ends_with(".jpg"))

#
# Only include samples whose `weather` field is "sunny"
#

view = dataset.match(F("weather").label == "sunny")

#
# Only include samples with at least 2 objects in their
# `predictions` field
#

view = dataset.match(F("predictions").detections.length() >= 2)

#
# Only include samples whose `predictions` field contains at least
# one object with area smaller than 0.2
#

# Bboxes are in [top-left-x, top-left-y, width, height] format
bbox = F("bounding_box")
bbox_area = bbox[2] * bbox[3]

small_boxes = F("predictions.detections").filter(bbox_area < 0.2)
view = dataset.match(small_boxes.length() > 0)

Parameters

filter –

a fiftyone.core.expressions.ViewExpression or MongoDB expression that returns a boolean describing the filter to apply

Returns

a fiftyone.core.view.DatasetView

match_frames(filter, omit_empty=True)¶

Filters the frames in the video collection by the given filter.

Examples:

import fiftyone as fo
import fiftyone.zoo as foz
from fiftyone import ViewField as F

dataset = foz.load_zoo_dataset("quickstart-video")

#
# Match frames with at least 10 detections
#

num_objects = F("detections.detections").length()
view = dataset.match_frames(num_objects > 10)

print(dataset.count())
print(view.count())

print(dataset.count("frames"))
print(view.count("frames"))

Parameters

filter –
a fiftyone.core.expressions.ViewExpression or MongoDB aggregation expression that returns a boolean describing the filter to apply
omit_empty (True) – whether to omit samples with no frame labels after filtering

Returns

a fiftyone.core.view.DatasetView

match_labels(labels=None, ids=None, tags=None, filter=None, fields=None, bool=None)¶

Selects the samples from the collection that contain (or do not contain) at least one label that matches the specified criteria.

Note that, unlike select_labels() and filter_labels(), this stage will not filter the labels themselves; it only selects the corresponding samples.

You can perform a selection via one or more of the following methods:

Provide the labels argument, which should contain a list of dicts in the format returned by fiftyone.core.session.Session.selected_labels, to match specific labels
Provide the ids argument to match labels with specific IDs
Provide the tags argument to match labels with specific tags
Provide the filter argument to match labels based on a boolean fiftyone.core.expressions.ViewExpression that is applied to each individual fiftyone.core.labels.Label element
Pass bool=False to negate the operation and instead match samples that do not contain at least one label matching the specified criteria

If multiple criteria are specified, labels must match all of them in order to trigger a sample match.

By default, the selection is applied to all fiftyone.core.labels.Label fields, but you can provide the fields argument to explicitly define the field(s) in which to search.

Examples:

import fiftyone as fo
import fiftyone.zoo as foz
from fiftyone import ViewField as F

dataset = foz.load_zoo_dataset("quickstart")

#
# Only show samples whose labels are currently selected in the App
#

session = fo.launch_app(dataset)

# Select some labels in the App...

view = dataset.match_labels(labels=session.selected_labels)

#
# Only include samples that contain labels with the specified IDs
#

# Grab some label IDs
ids = [
    dataset.first().ground_truth.detections[0].id,
    dataset.last().predictions.detections[0].id,
]

view = dataset.match_labels(ids=ids)

print(len(view))
print(view.count("ground_truth.detections"))
print(view.count("predictions.detections"))

#
# Only include samples that contain labels with the specified tags
#

# Grab some label IDs
ids = [
    dataset.first().ground_truth.detections[0].id,
    dataset.last().predictions.detections[0].id,
]

# Give the labels a "test" tag
dataset = dataset.clone()  # create copy since we're modifying data
dataset.select_labels(ids=ids).tag_labels("test")

print(dataset.count_values("ground_truth.detections.tags"))
print(dataset.count_values("predictions.detections.tags"))

# Retrieve the labels via their tag
view = dataset.match_labels(tags="test")

print(len(view))
print(view.count("ground_truth.detections"))
print(view.count("predictions.detections"))

#
# Only include samples that contain labels matching a filter
#

filter = F("confidence") > 0.99
view = dataset.match_labels(filter=filter, fields="predictions")

print(len(view))
print(view.count("ground_truth.detections"))
print(view.count("predictions.detections"))

Parameters

labels (None) – a list of dicts specifying the labels to select in the format returned by fiftyone.core.session.Session.selected_labels
ids (None) – an ID or iterable of IDs of the labels to select
tags (None) – a tag or iterable of tags of labels to select
filter (None) –
a fiftyone.core.expressions.ViewExpression or MongoDB aggregation expression that returns a boolean describing whether to select a given label. In the case of list fields like fiftyone.core.labels.Detections, the filter is applied to the list elements, not the root field
fields (None) – a field or iterable of fields from which to select
bool (None) – whether to match samples that have (None or True) or do not have (False) at least one label that matches the specified criteria

Returns

a fiftyone.core.view.DatasetView

match_tags(tags, bool=None, all=False)¶

Returns a view containing the samples in the collection that have or don’t have any/all of the given tag(s).

Examples:

import fiftyone as fo

dataset = fo.Dataset()
dataset.add_samples(
    [
        fo.Sample(filepath="image1.png", tags=["train"]),
        fo.Sample(filepath="image2.png", tags=["test"]),
        fo.Sample(filepath="image3.png", tags=["train", "test"]),
        fo.Sample(filepath="image4.png"),
    ]
)

#
# Only include samples that have the "test" tag
#

view = dataset.match_tags("test")

#
# Only include samples that do not have the "test" tag
#

view = dataset.match_tags("test", bool=False)

#
# Only include samples that have the "test" or "train" tags
#

view = dataset.match_tags(["test", "train"])

#
# Only include samples that have the "test" and "train" tags
#

view = dataset.match_tags(["test", "train"], all=True)

#
# Only include samples that do not have the "test" or "train" tags
#

view = dataset.match_tags(["test", "train"], bool=False)

#
# Only include samples that do not have the "test" and "train" tags
#

view = dataset.match_tags(["test", "train"], bool=False, all=True)

Parameters

tags – the tag or iterable of tags to match
bool (None) – whether to match samples that have (None or True) or do not have (False) the given tags
all (False) – whether to match samples that have (or don’t have) all (True) or any (False) of the given tags

Returns

a fiftyone.core.view.DatasetView

mean(field_or_expr, expr=None, safe=False)¶

Computes the arithmetic mean of the field values of the collection.

None-valued fields are ignored.

This aggregation is typically applied to numeric field types (or lists of such types):

Examples:

import fiftyone as fo
from fiftyone import ViewField as F

dataset = fo.Dataset()
dataset.add_samples(
    [
        fo.Sample(
            filepath="/path/to/image1.png",
            numeric_field=1.0,
            numeric_list_field=[1, 2, 3],
        ),
        fo.Sample(
            filepath="/path/to/image2.png",
            numeric_field=4.0,
            numeric_list_field=[1, 2],
        ),
        fo.Sample(
            filepath="/path/to/image3.png",
            numeric_field=None,
            numeric_list_field=None,
        ),
    ]
)

#
# Compute the mean of a numeric field
#

mean = dataset.mean("numeric_field")
print(mean)  # the mean

#
# Compute the mean of a numeric list field
#

mean = dataset.mean("numeric_list_field")
print(mean)  # the mean

#
# Compute the mean of a transformation of a numeric field
#

mean = dataset.mean(2 * (F("numeric_field") + 1))
print(mean)  # the mean

Parameters

field_or_expr –
a field name, embedded.field.name, fiftyone.core.expressions.ViewExpression, or MongoDB expression defining the field or expression to aggregate. This can also be a list or tuple of such arguments, in which case a tuple of corresponding aggregation results (each receiving the same additional keyword arguments, if any) will be returned
expr (None) –
a fiftyone.core.expressions.ViewExpression or MongoDB expression to apply to field_or_expr (which must be a field) before aggregating
safe (False) – whether to ignore nan/inf values when dealing with floating point values

Returns

the mean

merge_labels(in_field, out_field)¶

Merges the labels from the given input field into the given output field of the collection.

If this collection is a dataset, the input field is deleted after the merge.

If this collection is a view, the input field will still exist on the underlying dataset but will only contain the labels not present in this view.

Parameters

in_field – the name of the input label field
out_field – the name of the output label field, which will be created if necessary

mongo(pipeline, _needs_frames=None, _group_slices=None)¶

Adds a view stage defined by a raw MongoDB aggregation pipeline.

See MongoDB aggregation pipelines for more details.

Examples:

import fiftyone as fo

dataset = fo.Dataset()
dataset.add_samples(
    [
        fo.Sample(
            filepath="/path/to/image1.png",
            predictions=fo.Detections(
                detections=[
                    fo.Detection(
                        label="cat",
                        bounding_box=[0.1, 0.1, 0.5, 0.5],
                        confidence=0.9,
                    ),
                    fo.Detection(
                        label="dog",
                        bounding_box=[0.2, 0.2, 0.3, 0.3],
                        confidence=0.8,
                    ),
                ]
            ),
        ),
        fo.Sample(
            filepath="/path/to/image2.png",
            predictions=fo.Detections(
                detections=[
                    fo.Detection(
                        label="cat",
                        bounding_box=[0.5, 0.5, 0.4, 0.4],
                        confidence=0.95,
                    ),
                    fo.Detection(label="rabbit"),
                ]
            ),
        ),
        fo.Sample(
            filepath="/path/to/image3.png",
            predictions=fo.Detections(
                detections=[
                    fo.Detection(
                        label="squirrel",
                        bounding_box=[0.25, 0.25, 0.5, 0.5],
                        confidence=0.5,
                    ),
                ]
            ),
        ),
        fo.Sample(
            filepath="/path/to/image4.png",
            predictions=None,
        ),
    ]
)

#
# Extract a view containing the second and third samples in the
# dataset
#

view = dataset.mongo([{"$skip": 1}, {"$limit": 2}])

#
# Sort by the number of objects in the `precictions` field
#

view = dataset.mongo([
    {
        "$addFields": {
            "_sort_field": {
                "$size": {"$ifNull": ["$predictions.detections", []]}
            }
        }
    },
    {"$sort": {"_sort_field": -1}},
    {"$project": {"_sort_field": False}},
])

Parameters: pipeline – a MongoDB aggregation pipeline (list of dicts)
Returns: a fiftyone.core.view.DatasetView

quantiles(field_or_expr, quantiles, expr=None, safe=False)¶

Computes the quantile(s) of the field values of a collection.

None-valued fields are ignored.

This aggregation is typically applied to numeric field types (or lists of such types):

Examples:

import fiftyone as fo
from fiftyone import ViewField as F

dataset = fo.Dataset()
dataset.add_samples(
    [
        fo.Sample(
            filepath="/path/to/image1.png",
            numeric_field=1.0,
            numeric_list_field=[1, 2, 3],
        ),
        fo.Sample(
            filepath="/path/to/image2.png",
            numeric_field=4.0,
            numeric_list_field=[1, 2],
        ),
        fo.Sample(
            filepath="/path/to/image3.png",
            numeric_field=None,
            numeric_list_field=None,
        ),
    ]
)

#
# Compute the quantiles of a numeric field
#

quantiles = dataset.quantiles("numeric_field", [0.1, 0.5, 0.9])
print(quantiles)  # the quantiles

#
# Compute the quantiles of a numeric list field
#

quantiles = dataset.quantiles("numeric_list_field", [0.1, 0.5, 0.9])
print(quantiles)  # the quantiles

#
# Compute the mean of a transformation of a numeric field
#

quantiles = dataset.quantiles(2 * (F("numeric_field") + 1), [0.1, 0.5, 0.9])
print(quantiles)  # the quantiles

Parameters

field_or_expr –
a field name, embedded.field.name, fiftyone.core.expressions.ViewExpression, or MongoDB expression defining the field or expression to aggregate
quantiles – the quantile or iterable of quantiles to compute. Each quantile must be a numeric value in [0, 1]
expr (None) –
a fiftyone.core.expressions.ViewExpression or MongoDB expression to apply to field_or_expr (which must be a field) before aggregating
safe (False) – whether to ignore nan/inf values when dealing with floating point values

Returns

the quantile or list of quantiles

register_run(run_key, config, results=None, overwrite=False, cleanup=True, cache=True)¶

Registers a run under the given key on this collection.

Parameters

run_key – a run key
config – a fiftyone.core.runs.RunConfig
results (None) – an optional fiftyone.core.runs.RunResults
overwrite (False) – whether to allow overwriting an existing run of the same type
cleanup (True) – whether to execute an existing run’s fiftyone.core.runs.Run.cleanup() method when overwriting it
cache (True) – whether to cache the results on the collection

rename_annotation_run(anno_key, new_anno_key)¶

Replaces the key for the given annotation run with a new key.

Parameters

anno_key – an annotation key
new_anno_key – a new annotation key

rename_brain_run(brain_key, new_brain_key)¶

Replaces the key for the given brain run with a new key.

Parameters

brain_key – a brain key
new_brain_key – a new brain key

rename_evaluation(eval_key, new_eval_key)¶

Replaces the key for the given evaluation with a new key.

Parameters

eval_key – an evaluation key
new_anno_key – a new evaluation key

rename_run(run_key, new_run_key)¶

Replaces the key for the given run with a new key.

Parameters

run_key – a run key
new_run_key – a new run key

save_context(batch_size=None, batching_strategy=None)¶

Returns a context that can be used to save samples from this collection according to a configurable batching strategy.

Examples:

import random as r
import string as s

import fiftyone as fo
import fiftyone.zoo as foz

dataset = foz.load_zoo_dataset("cifar10", split="test")

def make_label():
    return "".join(r.choice(s.ascii_letters) for i in range(10))

# No save context
for sample in dataset.iter_samples(progress=True):
    sample.ground_truth.label = make_label()
    sample.save()

# Save using default batching strategy
with dataset.save_context() as context:
    for sample in dataset.iter_samples(progress=True):
        sample.ground_truth.label = make_label()
        context.save(sample)

# Save in batches of 10
with dataset.save_context(batch_size=10) as context:
    for sample in dataset.iter_samples(progress=True):
        sample.ground_truth.label = make_label()
        context.save(sample)

# Save every 0.5 seconds
with dataset.save_context(batch_size=0.5) as context:
    for sample in dataset.iter_samples(progress=True):
        sample.ground_truth.label = make_label()
        context.save(sample)

Parameters

batch_size (None) – the batch size to use. If a batching_strategy is provided, this parameter configures the strategy as described below. If no batching_strategy is provided, this can either be an integer specifying the number of samples to save in a batch (in which case batching_strategy is implicitly set to "static") or a float number of seconds between batched saves (in which case batching_strategy is implicitly set to "latency")
batching_strategy (None) –
the batching strategy to use for each save operation. Supported values are:
- "static": a fixed sample batch size for each save
- "size": a target batch size, in bytes, for each save
- "latency": a target latency, in seconds, between saves
By default, fo.config.default_batcher is used

Returns

a SaveContext

save_run_results(run_key, results, overwrite=True, cache=True)¶

Saves run results for the run with the given key.

Parameters

run_key – a run key
results – a fiftyone.core.runs.RunResults
overwrite (True) – whether to overwrite an existing result with the same key
cache (True) – whether to cache the results on the collection

schema(field_or_expr, expr=None, dynamic_only=False, _doc_type=None, _include_private=False)¶

Extracts the names and types of the attributes of a specified embedded document field across all samples in the collection.

Schema aggregations are useful for detecting the presence and types of dynamic attributes of fiftyone.core.labels.Label fields across a collection.

Examples:

import fiftyone as fo

dataset = fo.Dataset()

sample1 = fo.Sample(
    filepath="image1.png",
    ground_truth=fo.Detections(
        detections=[
            fo.Detection(
                label="cat",
                bounding_box=[0.1, 0.1, 0.4, 0.4],
                foo="bar",
                hello=True,
            ),
            fo.Detection(
                label="dog",
                bounding_box=[0.5, 0.5, 0.4, 0.4],
                hello=None,
            )
        ]
    )
)

sample2 = fo.Sample(
    filepath="image2.png",
    ground_truth=fo.Detections(
        detections=[
            fo.Detection(
                label="rabbit",
                bounding_box=[0.1, 0.1, 0.4, 0.4],
                foo=None,
            ),
            fo.Detection(
                label="squirrel",
                bounding_box=[0.5, 0.5, 0.4, 0.4],
                hello="there",
            ),
        ]
    )
)

dataset.add_samples([sample1, sample2])

#
# Get schema of all dynamic attributes on the detections in a
# `Detections` field
#

print(dataset.schema("ground_truth.detections", dynamic_only=True))
# {'foo': StringField, 'hello': [BooleanField, StringField]}

Parameters

field_or_expr –
a field name, embedded.field.name, fiftyone.core.expressions.ViewExpression, or MongoDB expression defining the field or expression to aggregate
expr (None) –
a fiftyone.core.expressions.ViewExpression or MongoDB expression to apply to field_or_expr (which must be a field) before aggregating
dynamic_only (False) – whether to only include dynamically added attributes

Returns

a dict mapping field names to fiftyone.core.fields.Field instances. If a field’s values takes multiple non-None types, the list of observed types will be returned

select(sample_ids, ordered=False)¶

Selects the samples with the given IDs from the collection.

Examples:

import fiftyone as fo
import fiftyone.zoo as foz

dataset = foz.load_zoo_dataset("quickstart")

#
# Create a view containing the currently selected samples in the App
#

session = fo.launch_app(dataset)

# Select samples in the App...

view = dataset.select(session.selected)

Parameters

sample_ids –

the samples to select. Can be any of the following:

a sample ID
an iterable of sample IDs
an iterable of booleans of same length as the collection encoding which samples to select
a fiftyone.core.sample.Sample or fiftyone.core.sample.SampleView
an iterable of fiftyone.core.sample.Sample or fiftyone.core.sample.SampleView instances
a fiftyone.core.collections.SampleCollection

ordered (False): whether to sort the samples in the returned view to: match the order of the provided IDs

Returns: a fiftyone.core.view.DatasetView

select_by(field, values, ordered=False)¶

Selects the samples with the given field values from the collection.

This stage is typically used to work with categorical fields (strings, ints, and bools). If you want to select samples based on floating point fields, use match().

Examples:

import fiftyone as fo

dataset = fo.Dataset()
dataset.add_samples(
    [
        fo.Sample(filepath="image%d.jpg" % i, int=i, str=str(i))
        for i in range(100)
    ]
)

#
# Create a view containing samples whose `int` field have the given
# values
#

view = dataset.select_by("int", [1, 51, 11, 41, 21, 31])
print(view.head(6))

#
# Create a view containing samples whose `str` field have the given
# values, in order
#

view = dataset.select_by(
    "str", ["1", "51", "11", "41", "21", "31"], ordered=True
)
print(view.head(6))

Parameters

field – a field or embedded.field.name
values – a value or iterable of values to select by
ordered (False) – whether to sort the samples in the returned view to match the order of the provided values

Returns

a fiftyone.core.view.DatasetView

select_fields(field_names=None, meta_filter=None, _allow_missing=False)¶

Selects only the fields with the given names from the samples in the collection. All other fields are excluded.

Note that default sample fields are always selected.

Examples:

import fiftyone as fo

dataset = fo.Dataset()
dataset.add_samples(
    [
        fo.Sample(
            filepath="/path/to/image1.png",
            uniqueness=1.0,
            ground_truth=fo.Detections(
                detections=[
                    fo.Detection(
                        label="cat",
                        bounding_box=[0.1, 0.1, 0.5, 0.5],
                        mood="surly",
                        age=51,
                    ),
                    fo.Detection(
                        label="dog",
                        bounding_box=[0.2, 0.2, 0.3, 0.3],
                        mood="happy",
                        age=52,
                    ),
                ]
            )
        ),
        fo.Sample(
            filepath="/path/to/image2.png",
            uniqueness=0.0,
        ),
        fo.Sample(
            filepath="/path/to/image3.png",
        ),
    ]
)

#
# Include only the default fields on each sample
#

view = dataset.select_fields()

#
# Include only the `uniqueness` field (and the default fields) on
# each sample
#

view = dataset.select_fields("uniqueness")

#
# Include only the `mood` attribute (and the default attributes) of
# each `Detection` in the `ground_truth` field
#

view = dataset.select_fields("ground_truth.detections.mood")

Parameters

field_names (None) – a field name or iterable of field names to select. May contain embedded.field.name as well
meta_filter (None) –
a filter that dynamically selects fields in the collection’s schema according to the specified rule, which can be matched against the field’s name, type, description, and/or info. For example:
- Use meta_filter="2023" or meta_filter={"any": "2023"} to select fields that have the string “2023” anywhere in their name, type, description, or info
- Use meta_filter={"type": "StringField"} or meta_filter={"type": "Classification"} to select all string or classification fields, respectively
- Use meta_filter={"description": "my description"} to select fields whose description contains the string “my description”
- Use meta_filter={"info": "2023"} to select fields that have the string “2023” anywhere in their info
- Use meta_filter={"info.key": "value"}} to select fields that have a specific key/value pair in their info
- Include meta_filter={"include_nested_fields": True, ...} in your meta filter to include all nested fields in the filter

Returns

a fiftyone.core.view.DatasetView

select_frames(frame_ids, omit_empty=True)¶

Selects the frames with the given IDs from the video collection.

Examples:

import fiftyone as fo
import fiftyone.zoo as foz

dataset = foz.load_zoo_dataset("quickstart-video")

#
# Select some specific frames
#

frame_ids = [
    dataset.first().frames.first().id,
    dataset.last().frames.last().id,
]

view = dataset.select_frames(frame_ids)

print(dataset.count())
print(view.count())

print(dataset.count("frames"))
print(view.count("frames"))

Parameters

frame_ids –
the frames to select. Can be any of the following:
- a frame ID
- an iterable of frame IDs
- a fiftyone.core.frame.Frame or fiftyone.core.frame.FrameView
- an iterable of fiftyone.core.frame.Frame or fiftyone.core.frame.FrameView instances
- a fiftyone.core.collections.SampleCollection whose frames to select
omit_empty (True) – whether to omit samples that have no frames after selecting the specified frames

Returns

a fiftyone.core.view.DatasetView

select_group_slices(slices=None, media_type=None, _allow_mixed=False, _force_mixed=False)¶

Selects the samples in the group collection from the given slice(s).

The returned view is a flattened non-grouped view containing only the slice(s) of interest.

Note

This stage performs a $lookup that pulls the requested slice(s) for each sample in the input collection from the source dataset. As a result, this stage always emits unfiltered samples.

Examples:

import fiftyone as fo

dataset = fo.Dataset()
dataset.add_group_field("group", default="ego")

group1 = fo.Group()
group2 = fo.Group()

dataset.add_samples(
    [
        fo.Sample(
            filepath="/path/to/left-image1.jpg",
            group=group1.element("left"),
        ),
        fo.Sample(
            filepath="/path/to/video1.mp4",
            group=group1.element("ego"),
        ),
        fo.Sample(
            filepath="/path/to/right-image1.jpg",
            group=group1.element("right"),
        ),
        fo.Sample(
            filepath="/path/to/left-image2.jpg",
            group=group2.element("left"),
        ),
        fo.Sample(
            filepath="/path/to/video2.mp4",
            group=group2.element("ego"),
        ),
        fo.Sample(
            filepath="/path/to/right-image2.jpg",
            group=group2.element("right"),
        ),
    ]
)

#
# Retrieve the samples from the "ego" group slice
#

view = dataset.select_group_slices("ego")

#
# Retrieve the samples from the "left" or "right" group slices
#

view = dataset.select_group_slices(["left", "right"])

#
# Retrieve all image samples
#

view = dataset.select_group_slices(media_type="image")

Parameters

slices (None) – a group slice or iterable of group slices to select. If neither argument is provided, a flattened list of all samples is returned
media_type (None) – a media type whose slice(s) to select

Returns

a fiftyone.core.view.DatasetView

select_groups(group_ids, ordered=False)¶

Selects the groups with the given IDs from the grouped collection.

Examples:

import fiftyone as fo
import fiftyone.zoo as foz

dataset = foz.load_zoo_dataset("quickstart-groups")

#
# Select some specific groups by ID
#

group_ids = dataset.take(10).values("group.id")

view = dataset.select_groups(group_ids)

assert set(view.values("group.id")) == set(group_ids)

view = dataset.select_groups(group_ids, ordered=True)

assert view.values("group.id") == group_ids

Parameters

groups_ids –
the groups to select. Can be any of the following:
- a group ID
- an iterable of group IDs
- a fiftyone.core.sample.Sample or fiftyone.core.sample.SampleView
- a group dict returned by get_group()
- an iterable of fiftyone.core.sample.Sample or fiftyone.core.sample.SampleView instances
- an iterable of group dicts returned by get_group()
- a fiftyone.core.collections.SampleCollection
ordered (False) – whether to sort the groups in the returned view to match the order of the provided IDs

Returns

a fiftyone.core.view.DatasetView

select_labels(labels=None, ids=None, tags=None, fields=None, omit_empty=True)¶

Selects only the specified labels from the collection.

The returned view will omit samples, sample fields, and individual labels that do not match the specified selection criteria.

You can perform a selection via one or more of the following methods:

Provide the labels argument, which should contain a list of dicts in the format returned by fiftyone.core.session.Session.selected_labels, to select specific labels
Provide the ids argument to select labels with specific IDs
Provide the tags argument to select labels with specific tags

If multiple criteria are specified, labels must match all of them in order to be selected.

By default, the selection is applied to all fiftyone.core.labels.Label fields, but you can provide the fields argument to explicitly define the field(s) in which to select.

Examples:

import fiftyone as fo
import fiftyone.zoo as foz

dataset = foz.load_zoo_dataset("quickstart")

#
# Only include the labels currently selected in the App
#

session = fo.launch_app(dataset)

# Select some labels in the App...

view = dataset.select_labels(labels=session.selected_labels)

#
# Only include labels with the specified IDs
#

# Grab some label IDs
ids = [
    dataset.first().ground_truth.detections[0].id,
    dataset.last().predictions.detections[0].id,
]

view = dataset.select_labels(ids=ids)

print(view.count("ground_truth.detections"))
print(view.count("predictions.detections"))

#
# Only include labels with the specified tags
#

# Grab some label IDs
ids = [
    dataset.first().ground_truth.detections[0].id,
    dataset.last().predictions.detections[0].id,
]

# Give the labels a "test" tag
dataset = dataset.clone()  # create copy since we're modifying data
dataset.select_labels(ids=ids).tag_labels("test")

print(dataset.count_label_tags())

# Retrieve the labels via their tag
view = dataset.select_labels(tags="test")

print(view.count("ground_truth.detections"))
print(view.count("predictions.detections"))

Parameters

labels (None) – a list of dicts specifying the labels to select in the format returned by fiftyone.core.session.Session.selected_labels
ids (None) – an ID or iterable of IDs of the labels to select
tags (None) – a tag or iterable of tags of labels to select
fields (None) – a field or iterable of fields from which to select
omit_empty (True) – whether to omit samples that have no labels after filtering

Returns

a fiftyone.core.view.DatasetView

set_field(field, expr, _allow_missing=False)¶

Sets a field or embedded field on each sample in a collection by evaluating the given expression.

This method can process embedded list fields. To do so, simply append [] to any list component(s) of the field path.

Note

There are two cases where FiftyOne will automatically unwind array fields without requiring you to explicitly specify this via the [] syntax:

Top-level lists: when you specify a field path that refers to a top-level list field of a dataset; i.e., list_field is automatically coerced to list_field[], if necessary.

List fields: When you specify a field path that refers to the list field of a Label class, such as the Detections.detections attribute; i.e., ground_truth.detections.label is automatically coerced to ground_truth.detections[].label, if necessary.

See the examples below for demonstrations of this behavior.

The provided expr is interpreted relative to the document on which the embedded field is being set. For example, if you are setting a nested field field="embedded.document.field", then the expression expr you provide will be applied to the embedded.document document. Note that you can override this behavior by defining an expression that is bound to the root document by prepending "$" to any field name(s) in the expression.

See the examples below for more information.

Note

Note that you cannot set a non-existing top-level field using this stage, since doing so would violate the dataset’s schema. You can, however, first declare a new field via fiftyone.core.dataset.Dataset.add_sample_field() and then populate it in a view via this stage.

Examples:

import fiftyone as fo
import fiftyone.zoo as foz
from fiftyone import ViewField as F

dataset = foz.load_zoo_dataset("quickstart")

#
# Replace all values of the `uniqueness` field that are less than
# 0.5 with `None`
#

view = dataset.set_field(
    "uniqueness",
    (F("uniqueness") >= 0.5).if_else(F("uniqueness"), None)
)
print(view.bounds("uniqueness"))

#
# Lower bound all object confidences in the `predictions` field at
# 0.5
#

view = dataset.set_field(
    "predictions.detections.confidence", F("confidence").max(0.5)
)
print(view.bounds("predictions.detections.confidence"))

#
# Add a `num_predictions` property to the `predictions` field that
# contains the number of objects in the field
#

view = dataset.set_field(
    "predictions.num_predictions",
    F("$predictions.detections").length(),
)
print(view.bounds("predictions.num_predictions"))

#
# Set an `is_animal` field on each object in the `predictions` field
# that indicates whether the object is an animal
#

ANIMALS = [
    "bear", "bird", "cat", "cow", "dog", "elephant", "giraffe",
    "horse", "sheep", "zebra"
]

view = dataset.set_field(
    "predictions.detections.is_animal", F("label").is_in(ANIMALS)
)
print(view.count_values("predictions.detections.is_animal"))

Parameters

field – the field or embedded.field.name to set
expr –
a fiftyone.core.expressions.ViewExpression or MongoDB expression that defines the field value to set

Returns

a fiftyone.core.view.DatasetView

set_label_values(field_name, values, dynamic=False, skip_none=False, validate=True, progress=False)¶

Sets the fields of the specified labels in the collection to the given values.

Note

This method is appropriate when you have the IDs of the labels you wish to modify. See set_values() and set_field() if your updates are not keyed by label ID.

Examples:

import fiftyone as fo
import fiftyone.zoo as foz
from fiftyone import ViewField as F

dataset = foz.load_zoo_dataset("quickstart")

#
# Populate a new boolean attribute on all high confidence labels
#

view = dataset.filter_labels("predictions", F("confidence") > 0.99)

label_ids = view.values("predictions.detections.id", unwind=True)
values = {_id: True for _id in label_ids}

dataset.set_label_values("predictions.detections.high_conf", values)

print(dataset.count("predictions.detections"))
print(len(label_ids))
print(dataset.count_values("predictions.detections.high_conf"))

Parameters

field_name – a field or embedded.field.name
values – a dict mapping label IDs to values
skip_none (False) – whether to treat None data in values as missing data that should not be set
dynamic (False) – whether to declare dynamic attributes of embedded document fields that are encountered
validate (True) – whether to validate that the values are compliant with the dataset schema before adding them
progress (False) – whether to render a progress bar (True/False), use the default value fiftyone.config.show_progress_bars (None), or a progress callback function to invoke instead

set_values(field_name, values, key_field=None, skip_none=False, expand_schema=True, dynamic=False, validate=True, progress=False, _allow_missing=False, _sample_ids=None, _frame_ids=None)¶

Sets the field or embedded field on each sample or frame in the collection to the given values.

When setting a sample field embedded.field.name, this function is an efficient implementation of the following loop:

for sample, value in zip(sample_collection, values):
    sample.embedded.field.name = value
    sample.save()

When setting an embedded field that contains an array, say embedded.array.field.name, this function is an efficient implementation of the following loop:

for sample, array_values in zip(sample_collection, values):
    for doc, value in zip(sample.embedded.array, array_values):
        doc.field.name = value

    sample.save()

When setting a frame field frames.embedded.field.name, this function is an efficient implementation of the following loop:

for sample, frame_values in zip(sample_collection, values):
    for frame, value in zip(sample.frames.values(), frame_values):
        frame.embedded.field.name = value

    sample.save()

When setting an embedded frame field that contains an array, say frames.embedded.array.field.name, this function is an efficient implementation of the following loop:

for sample, frame_values in zip(sample_collection, values):
    for frame, array_values in zip(sample.frames.values(), frame_values):
        for doc, value in zip(frame.embedded.array, array_values):
            doc.field.name = value

    sample.save()

When values is a dict mapping keys in key_field to values, then this function is an efficient implementation of the following loop:

for key, value in values.items():
    sample = sample_collection.one(F(key_field) == key)
    sample.embedded.field.name = value
    sample.save()

When setting frame fields using the dict values syntax, each value in values may either be a list corresponding to the frames of the sample matching the given key, or each value may itself be a dict mapping frame numbers to values. In the latter case, this function is an efficient implementation of the following loop:

for key, frame_values in values.items():
    sample = sample_collection.one(F(key_field) == key)
    for frame_number, value in frame_values.items():
        frame = sample[frame_number]
        frame.embedded.field.name = value

    sample.save()

You can also update list fields using the dict values syntax, in which case this method is an efficient implementation of the natural nested list modifications of the above sample/frame loops.

The dual function of set_values() is values(), which can be used to efficiently extract the values of a field or embedded field of all samples in a collection as lists of values in the same structure expected by this method.

Note

If the values you are setting can be described by a fiftyone.core.expressions.ViewExpression applied to the existing dataset contents, then consider using set_field() + save() for an even more efficient alternative to explicitly iterating over the dataset or calling values() + set_values() to perform the update in-memory.

Examples:

import random

import fiftyone as fo
import fiftyone.zoo as foz
from fiftyone import ViewField as F

dataset = foz.load_zoo_dataset("quickstart")

#
# Create a new sample field
#

values = [random.random() for _ in range(len(dataset))]
dataset.set_values("random", values)

print(dataset.bounds("random"))

#
# Add a tag to all low confidence labels
#

view = dataset.filter_labels("predictions", F("confidence") < 0.06)

detections = view.values("predictions.detections")
for sample_detections in detections:
    for detection in sample_detections:
        detection.tags.append("low_confidence")

view.set_values("predictions.detections", detections)

print(dataset.count_label_tags())

Parameters

field_name – a field or embedded.field.name
values – an iterable of values, one for each sample in the collection. When setting frame fields, each element can either be an iterable of values (one for each existing frame of the sample) or a dict mapping frame numbers to values. If field_name contains array fields, the corresponding elements of values must be arrays of the same lengths. This argument can also be a dict mapping keys to values (each value as described previously), in which case the keys are used to match samples by their key_field
key_field (None) – a key field to use when choosing which samples to update when values is a dict
skip_none (False) – whether to treat None data in values as missing data that should not be set
expand_schema (True) – whether to dynamically add new sample/frame fields encountered to the dataset schema. If False, an error is raised if the root field_name does not exist
dynamic (False) – whether to declare dynamic attributes of embedded document fields that are encountered
validate (True) – whether to validate that the values are compliant with the dataset schema before adding them
progress (False) – whether to render a progress bar (True/False), use the default value fiftyone.config.show_progress_bars (None), or a progress callback function to invoke instead

shuffle(seed=None)¶

Randomly shuffles the samples in the collection.

Examples:

import fiftyone as fo

dataset = fo.Dataset()
dataset.add_samples(
    [
        fo.Sample(
            filepath="/path/to/image1.png",
            ground_truth=fo.Classification(label="cat"),
        ),
        fo.Sample(
            filepath="/path/to/image2.png",
            ground_truth=fo.Classification(label="dog"),
        ),
        fo.Sample(
            filepath="/path/to/image3.png",
            ground_truth=None,
        ),
    ]
)

#
# Return a view that contains a randomly shuffled version of the
# samples in the dataset
#

view = dataset.shuffle()

#
# Shuffle the samples with a fixed random seed
#

view = dataset.shuffle(seed=51)

Parameters: seed (None) – an optional random seed to use when shuffling the samples
Returns: a fiftyone.core.view.DatasetView

skip(skip)¶

Omits the given number of samples from the head of the collection.

Examples:

import fiftyone as fo

dataset = fo.Dataset()
dataset.add_samples(
    [
        fo.Sample(
            filepath="/path/to/image1.png",
            ground_truth=fo.Classification(label="cat"),
        ),
        fo.Sample(
            filepath="/path/to/image2.png",
            ground_truth=fo.Classification(label="dog"),
        ),
        fo.Sample(
            filepath="/path/to/image3.png",
            ground_truth=fo.Classification(label="rabbit"),
        ),
        fo.Sample(
            filepath="/path/to/image4.png",
            ground_truth=None,
        ),
    ]
)

#
# Omit the first two samples from the dataset
#

view = dataset.skip(2)

Parameters: skip – the number of samples to skip. If a non-positive number is provided, no samples are omitted
Returns: a fiftyone.core.view.DatasetView

sort_by(field_or_expr, reverse=False, create_index=True)¶

Sorts the samples in the collection by the given field(s) or expression(s).

Examples:

import fiftyone as fo
import fiftyone.zoo as foz
from fiftyone import ViewField as F

dataset = foz.load_zoo_dataset("quickstart")

#
# Sort the samples by their `uniqueness` field in ascending order
#

view = dataset.sort_by("uniqueness", reverse=False)

#
# Sorts the samples in descending order by the number of detections
# in their `predictions` field whose bounding box area is less than
# 0.2
#

# Bboxes are in [top-left-x, top-left-y, width, height] format
bbox = F("bounding_box")
bbox_area = bbox[2] * bbox[3]

small_boxes = F("predictions.detections").filter(bbox_area < 0.2)
view = dataset.sort_by(small_boxes.length(), reverse=True)

#
# Performs a compound sort where samples are first sorted in
# descending or by number of detections and then in ascending order
# of uniqueness for samples with the same number of predictions
#

view = dataset.sort_by(
    [
        (F("predictions.detections").length(), -1),
        ("uniqueness", 1),
    ]
)

num_objects, uniqueness = view[:5].values(
    [F("predictions.detections").length(), "uniqueness"]
)
print(list(zip(num_objects, uniqueness)))

Parameters

field_or_expr –
the field(s) or expression(s) to sort by. This can be any of the following:
- a field to sort by
- an embedded.field.name to sort by
- a fiftyone.core.expressions.ViewExpression or a MongoDB aggregation expression that defines the quantity to sort by
- a list of (field_or_expr, order) tuples defining a compound sort criteria, where field_or_expr is a field or expression as defined above, and order can be 1 or any string starting with “a” for ascending order, or -1 or any string starting with “d” for descending order
reverse (False) – whether to return the results in descending order
create_index (True) – whether to create an index, if necessary, to optimize the sort. Only applicable when sorting by field(s), not expressions

Returns

a fiftyone.core.view.DatasetView

sort_by_similarity(query, k=None, reverse=False, dist_field=None, brain_key=None)¶

Sorts the collection by similarity to a specified query.

In order to use this stage, you must first use fiftyone.brain.compute_similarity() to index your dataset by similarity.

Examples:

import fiftyone as fo
import fiftyone.brain as fob
import fiftyone.zoo as foz

dataset = foz.load_zoo_dataset("quickstart")

fob.compute_similarity(
    dataset, model="clip-vit-base32-torch", brain_key="clip"
)

#
# Sort samples by their similarity to a sample by its ID
#

query_id = dataset.first().id

view = dataset.sort_by_similarity(query_id, k=5)

#
# Sort samples by their similarity to a manually computed vector
#

model = foz.load_zoo_model("clip-vit-base32-torch")
embeddings = dataset.take(2, seed=51).compute_embeddings(model)
query = embeddings.mean(axis=0)

view = dataset.sort_by_similarity(query, k=5)

#
# Sort samples by their similarity to a text prompt
#

query = "kites high in the air"

view = dataset.sort_by_similarity(query, k=5)

Parameters

query –
the query, which can be any of the following:
- an ID or iterable of IDs
- a num_dims vector or num_queries x num_dims array of vectors
- a prompt or iterable of prompts (if supported by the index)
k (None) – the number of matches to return. By default, the entire collection is sorted
reverse (False) – whether to sort by least similarity (True) or greatest similarity (False). Some backends may not support least similarity
dist_field (None) – the name of a float field in which to store the distance of each example to the specified query. The field is created if necessary
brain_key (None) – the brain key of an existing fiftyone.brain.compute_similarity() run on the dataset. If not specified, the dataset must have an applicable run, which will be used by default

Returns

a fiftyone.core.view.DatasetView

split_labels(in_field, out_field, filter=None)¶

Splits the labels from the given input field into the given output field of the collection.

This method is typically invoked on a view that has filtered the contents of the specified input field, so that the labels in the view are moved to the output field and the remaining labels are left in-place.

Alternatively, you can provide a filter expression that selects the labels of interest to move in this collection.

Parameters

in_field – the name of the input label field
out_field – the name of the output label field, which will be created if necessary
filter (None) – a boolean fiftyone.core.expressions.ViewExpression to apply to each label in the input field to determine whether to move it (True) or leave it (False)

std(field_or_expr, expr=None, safe=False, sample=False)¶

Computes the standard deviation of the field values of the collection.

None-valued fields are ignored.

This aggregation is typically applied to numeric field types (or lists of such types):

Examples:

import fiftyone as fo
from fiftyone import ViewField as F

dataset = fo.Dataset()
dataset.add_samples(
    [
        fo.Sample(
            filepath="/path/to/image1.png",
            numeric_field=1.0,
            numeric_list_field=[1, 2, 3],
        ),
        fo.Sample(
            filepath="/path/to/image2.png",
            numeric_field=4.0,
            numeric_list_field=[1, 2],
        ),
        fo.Sample(
            filepath="/path/to/image3.png",
            numeric_field=None,
            numeric_list_field=None,
        ),
    ]
)

#
# Compute the standard deviation of a numeric field
#

std = dataset.std("numeric_field")
print(std)  # the standard deviation

#
# Compute the standard deviation of a numeric list field
#

std = dataset.std("numeric_list_field")
print(std)  # the standard deviation

#
# Compute the standard deviation of a transformation of a numeric field
#

std = dataset.std(2 * (F("numeric_field") + 1))
print(std)  # the standard deviation

Parameters

field_or_expr –
a field name, embedded.field.name, fiftyone.core.expressions.ViewExpression, or MongoDB expression defining the field or expression to aggregate. This can also be a list or tuple of such arguments, in which case a tuple of corresponding aggregation results (each receiving the same additional keyword arguments, if any) will be returned
expr (None) –
a fiftyone.core.expressions.ViewExpression or MongoDB expression to apply to field_or_expr (which must be a field) before aggregating
safe (False) – whether to ignore nan/inf values when dealing with floating point values
sample (False) – whether to compute the sample standard deviation rather than the population standard deviation

Returns

the standard deviation

sum(field_or_expr, expr=None, safe=False)¶

Computes the sum of the field values of the collection.

None-valued fields are ignored.

This aggregation is typically applied to numeric field types (or lists of such types):

Examples:

import fiftyone as fo
from fiftyone import ViewField as F

dataset = fo.Dataset()
dataset.add_samples(
    [
        fo.Sample(
            filepath="/path/to/image1.png",
            numeric_field=1.0,
            numeric_list_field=[1, 2, 3],
        ),
        fo.Sample(
            filepath="/path/to/image2.png",
            numeric_field=4.0,
            numeric_list_field=[1, 2],
        ),
        fo.Sample(
            filepath="/path/to/image3.png",
            numeric_field=None,
            numeric_list_field=None,
        ),
    ]
)

#
# Compute the sum of a numeric field
#

total = dataset.sum("numeric_field")
print(total)  # the sum

#
# Compute the sum of a numeric list field
#

total = dataset.sum("numeric_list_field")
print(total)  # the sum

#
# Compute the sum of a transformation of a numeric field
#

total = dataset.sum(2 * (F("numeric_field") + 1))
print(total)  # the sum

Parameters

field_or_expr –
a field name, embedded.field.name, fiftyone.core.expressions.ViewExpression, or MongoDB expression defining the field or expression to aggregate. This can also be a list or tuple of such arguments, in which case a tuple of corresponding aggregation results (each receiving the same additional keyword arguments, if any) will be returned
expr (None) –
a fiftyone.core.expressions.ViewExpression or MongoDB expression to apply to field_or_expr (which must be a field) before aggregating
safe (False) – whether to ignore nan/inf values when dealing with floating point values

Returns

the sum

sync_last_modified_at(include_frames=True)¶

Syncs the last_modified_at property(s) of the dataset.

Updates the last_modified_at property of the dataset if necessary to incorporate any modification timestamps to its samples.

If include_frames==True, the last_modified_at property of each video sample is first updated if necessary to incorporate any modification timestamps to its frames.

Parameters: include_frames (True) – whether to update the last_modified_at property of video samples. Only applicable to datasets that contain videos

tag_labels(tags, label_fields=None)¶

Adds the tag(s) to all labels in the specified label field(s) of this collection, if necessary.

Parameters

tags – a tag or iterable of tags
label_fields (None) – an optional name or iterable of names of fiftyone.core.labels.Label fields. By default, all label fields are used

tag_samples(tags)¶

Adds the tag(s) to all samples in this collection, if necessary.

Parameters: tags – a tag or iterable of tags

take(size, seed=None)¶

Randomly samples the given number of samples from the collection.

Examples:

import fiftyone as fo

dataset = fo.Dataset()
dataset.add_samples(
    [
        fo.Sample(
            filepath="/path/to/image1.png",
            ground_truth=fo.Classification(label="cat"),
        ),
        fo.Sample(
            filepath="/path/to/image2.png",
            ground_truth=fo.Classification(label="dog"),
        ),
        fo.Sample(
            filepath="/path/to/image3.png",
            ground_truth=fo.Classification(label="rabbit"),
        ),
        fo.Sample(
            filepath="/path/to/image4.png",
            ground_truth=None,
        ),
    ]
)

#
# Take two random samples from the dataset
#

view = dataset.take(2)

#
# Take two random samples from the dataset with a fixed seed
#

view = dataset.take(2, seed=51)

Parameters

size – the number of samples to return. If a non-positive number is provided, an empty view is returned
seed (None) – an optional random seed to use when selecting the samples

Returns

a fiftyone.core.view.DatasetView

to_clips(field_or_expr, **kwargs)¶

Creates a view that contains one sample per clip defined by the given field or expression in the video collection.

The returned view will contain:

A sample_id field that records the sample ID from which each clip was taken
A support field that records the [first, last] frame support of each clip
All frame-level information from the underlying dataset of the input collection

Refer to fiftyone.core.clips.make_clips_dataset() to see the available configuration options for generating clips.

Note

The clip generation logic will respect any frame-level modifications defined in the input collection, but the output clips will always contain all frame-level labels.

Examples:

import fiftyone as fo
import fiftyone.zoo as foz
from fiftyone import ViewField as F

dataset = foz.load_zoo_dataset("quickstart-video")

#
# Create a clips view that contains one clip for each contiguous
# segment that contains at least one road sign in every frame
#

clips = (
    dataset
    .filter_labels("frames.detections", F("label") == "road sign")
    .to_clips("frames.detections")
)
print(clips)

#
# Create a clips view that contains one clip for each contiguous
# segment that contains at least two road signs in every frame
#

signs = F("detections.detections").filter(F("label") == "road sign")
clips = dataset.to_clips(signs.length() >= 2)
print(clips)

Parameters

field_or_expr –
can be any of the following:
- a fiftyone.core.labels.TemporalDetection, fiftyone.core.labels.TemporalDetections, fiftyone.core.fields.FrameSupportField, or list of fiftyone.core.fields.FrameSupportField field
- a frame-level label list field of any of the following types:
- a fiftyone.core.expressions.ViewExpression that returns a boolean to apply to each frame of the input collection to determine if the frame should be clipped
- a list of [(first1, last1), (first2, last2), ...] lists defining the frame numbers of the clips to extract from each sample
other_fields (None) –
controls whether sample fields other than the default sample fields are included. Can be any of the following:
- a field or list of fields to include
- True to include all other fields
- None/False to include no other fields
tol (0) – the maximum number of false frames that can be overlooked when generating clips. Only applicable when field_or_expr is a frame-level list field or expression
min_len (0) – the minimum allowable length of a clip, in frames. Only applicable when field_or_expr is a frame-level list field or an expression
trajectories (False) – whether to create clips for each unique object trajectory defined by their (label, index). Only applicable when field_or_expr is a frame-level field

Returns

a fiftyone.core.clips.ClipsView

to_dict(rel_dir=None, include_private=False, include_frames=False, frame_labels_dir=None, pretty_print=False, progress=None)¶

Returns a JSON dictionary representation of the collection.

Parameters

rel_dir (None) – a relative directory to remove from the filepath of each sample, if possible. The path is converted to an absolute path (if necessary) via fiftyone.core.storage.normalize_path(). The typical use case for this argument is that your source data lives in a single directory and you wish to serialize relative, rather than absolute, paths to the data within that directory
include_private (False) – whether to include private fields
include_frames (False) – whether to include the frame labels for video samples
frame_labels_dir (None) – a directory in which to write per-sample JSON files containing the frame labels for video samples. If omitted, frame labels will be included directly in the returned JSON dict (which can be quite quite large for video datasets containing many frames). Only applicable to datasets that contain videos when include_frames is True
pretty_print (False) – whether to render frame labels JSON in human readable format with newlines and indentations. Only applicable to datasets that contain videos when a frame_labels_dir is provided
progress (None) – whether to render a progress bar (True/False), use the default value fiftyone.config.show_progress_bars (None), or a progress callback function to invoke instead

Returns

a JSON dict

to_evaluation_patches(eval_key, **kwargs)¶

Creates a view based on the results of the evaluation with the given key that contains one sample for each true positive, false positive, and false negative example in the collection, respectively.

True positive examples will result in samples with both their ground truth and predicted fields populated, while false positive/negative examples will only have one of their corresponding predicted/ground truth fields populated, respectively.

If multiple predictions are matched to a ground truth object (e.g., if the evaluation protocol includes a crowd attribute), then all matched predictions will be stored in the single sample along with the ground truth object.

The returned dataset will also have top-level type and iou fields populated based on the evaluation results for that example, as well as a sample_id field recording the sample ID of the example, and a crowd field if the evaluation protocol defines a crowd attribute.

Note

The returned view will contain patches for the contents of this collection, which may differ from the view on which the eval_key evaluation was performed. This may exclude some labels that were evaluated and/or include labels that were not evaluated.

If you would like to see patches for the exact view on which an evaluation was performed, first call load_evaluation_view() to load the view and then convert to patches.

Examples:

import fiftyone as fo
import fiftyone.zoo as foz

dataset = foz.load_zoo_dataset("quickstart")
dataset.evaluate_detections("predictions", eval_key="eval")

session = fo.launch_app(dataset)

#
# Create a patches view for the evaluation results
#

view = dataset.to_evaluation_patches("eval")
print(view)

session.view = view

Parameters

eval_key – an evaluation key that corresponds to the evaluation of ground truth/predicted fields that are of type fiftyone.core.labels.Detections, fiftyone.core.labels.Polylines, or fiftyone.core.labels.Keypoints
other_fields (None) –
controls whether fields other than the ground truth/predicted fields and the default sample fields are included. Can be any of the following:
- a field or list of fields to include
- True to include all other fields
- None/False to include no other fields

Returns

a fiftyone.core.patches.EvaluationPatchesView

to_frames(**kwargs)¶

Creates a view that contains one sample per frame in the video collection.

The returned view will contain all frame-level fields and the tags of each video as sample-level fields, as well as a sample_id field that records the IDs of the parent sample for each frame.

By default, sample_frames is False and this method assumes that the frames of the input collection have filepath fields populated pointing to each frame image. Any frames without a filepath populated will be omitted from the returned view.

When sample_frames is True, this method samples each video in the collection into a directory of per-frame images and stores the filepaths in the filepath frame field of the source dataset. By default, each folder of images is written using the same basename as the input video. For example, if frames_patt = "%%06d.jpg", then videos with the following paths:

/path/to/video1.mp4
/path/to/video2.mp4
...

would be sampled as follows:

/path/to/video1/
    000001.jpg
    000002.jpg
    ...
/path/to/video2/
    000001.jpg
    000002.jpg
    ...

However, you can use the optional output_dir and rel_dir parameters to customize the location and shape of the sampled frame folders. For example, if output_dir = "/tmp" and rel_dir = "/path/to", then videos with the following paths:

/path/to/folderA/video1.mp4
/path/to/folderA/video2.mp4
/path/to/folderB/video3.mp4
...

would be sampled as follows:

/tmp/folderA/
    video1/
        000001.jpg
        000002.jpg
        ...
    video2/
        000001.jpg
        000002.jpg
        ...
/tmp/folderB/
    video3/
        000001.jpg
        000002.jpg
        ...

By default, samples will be generated for every video frame at full resolution, but this method provides a variety of parameters that can be used to customize the sampling behavior.

Note

If this method is run multiple times with sample_frames set to True, existing frames will not be resampled unless you set force_sample to True.

Examples:

import fiftyone as fo
import fiftyone.zoo as foz
from fiftyone import ViewField as F

dataset = foz.load_zoo_dataset("quickstart-video")

session = fo.launch_app(dataset)

#
# Create a frames view for an entire video dataset
#

frames = dataset.to_frames(sample_frames=True)
print(frames)

session.view = frames

#
# Create a frames view that only contains frames with at least 10
# objects, sampled at a maximum frame rate of 1fps
#

num_objects = F("detections.detections").length()
view = dataset.match_frames(num_objects > 10)

frames = view.to_frames(max_fps=1)
print(frames)

session.view = frames

Parameters

sample_frames (False) – whether to assume that the frame images have already been sampled at locations stored in the filepath field of each frame (False), or whether to sample the video frames now according to the specified parameters (True)
fps (None) – an optional frame rate at which to sample each video’s frames
max_fps (None) – an optional maximum frame rate at which to sample. Videos with frame rate exceeding this value are downsampled
size (None) – an optional (width, height) at which to sample frames. A dimension can be -1, in which case the aspect ratio is preserved. Only applicable when sample_frames=True
min_size (None) – an optional minimum (width, height) for each frame. A dimension can be -1 if no constraint should be applied. The frames are resized (aspect-preserving) if necessary to meet this constraint. Only applicable when sample_frames=True
max_size (None) – an optional maximum (width, height) for each frame. A dimension can be -1 if no constraint should be applied. The frames are resized (aspect-preserving) if necessary to meet this constraint. Only applicable when sample_frames=True
sparse (False) – whether to only sample frame images for frame numbers for which fiftyone.core.frame.Frame instances exist in the input collection. This parameter has no effect when sample_frames==False since frames must always exist in order to have filepath information use
output_dir (None) – an optional output directory in which to write the sampled frames. By default, the frames are written in folders with the same basename of each video
rel_dir (None) – a relative directory to remove from the filepath of each video, if possible. The path is converted to an absolute path (if necessary) via fiftyone.core.storage.normalize_path(). This argument can be used in conjunction with output_dir to cause the sampled frames to be written in a nested directory structure within output_dir matching the shape of the input video’s folder structure
frames_patt (None) – a pattern specifying the filename/format to use to write or check or existing sampled frames, e.g., "%%06d.jpg". The default value is fiftyone.config.default_sequence_idx + fiftyone.config.default_image_ext
force_sample (False) – whether to resample videos whose sampled frames already exist. Only applicable when sample_frames=True
skip_failures (True) – whether to gracefully continue without raising an error if a video cannot be sampled
verbose (False) – whether to log information about the frames that will be sampled, if any

Returns

a fiftyone.core.video.FramesView

to_json(rel_dir=None, include_private=False, include_frames=False, frame_labels_dir=None, pretty_print=False)¶

Returns a JSON string representation of the collection.

The samples will be written as a list in a top-level samples field of the returned dictionary.

Parameters

rel_dir (None) – a relative directory to remove from the filepath of each sample, if possible. The path is converted to an absolute path (if necessary) via fiftyone.core.storage.normalize_path(). The typical use case for this argument is that your source data lives in a single directory and you wish to serialize relative, rather than absolute, paths to the data within that directory
include_private (False) – whether to include private fields
include_frames (False) – whether to include the frame labels for video samples
frame_labels_dir (None) – a directory in which to write per-sample JSON files containing the frame labels for video samples. If omitted, frame labels will be included directly in the returned JSON dict (which can be quite quite large for video datasets containing many frames). Only applicable to datasets that contain videos when include_frames is True
pretty_print (False) – whether to render the JSON in human readable format with newlines and indentations

Returns

a JSON string

to_patches(field, **kwargs)¶

Creates a view that contains one sample per object patch in the specified field of the collection.

Fields other than field and the default sample fields will not be included in the returned view. A sample_id field will be added that records the sample ID from which each patch was taken.

Examples:

import fiftyone as fo
import fiftyone.zoo as foz

dataset = foz.load_zoo_dataset("quickstart")

session = fo.launch_app(dataset)

#
# Create a view containing the ground truth patches
#

view = dataset.to_patches("ground_truth")
print(view)

session.view = view

Parameters

field – the patches field, which must be of type fiftyone.core.labels.Detections, fiftyone.core.labels.Polylines, or fiftyone.core.labels.Keypoints
other_fields (None) –
controls whether fields other than field and the default sample fields are included. Can be any of the following:
- a field or list of fields to include
- True to include all other fields
- None/False to include no other fields
keep_label_lists (False) – whether to store the patches in label list fields of the same type as the input collection rather than using their single label variants

Returns

a fiftyone.core.patches.PatchesView

to_trajectories(field, **kwargs)¶

Creates a view that contains one clip for each unique object trajectory defined by their (label, index) in a frame-level field of a video collection.

The returned view will contain:

A sample_id field that records the sample ID from which each clip was taken
A support field that records the [first, last] frame support of each clip
A sample-level label field that records the label and index of each trajectory

Examples:

import fiftyone as fo
import fiftyone.zoo as foz
from fiftyone import ViewField as F

dataset = foz.load_zoo_dataset("quickstart-video")

#
# Create a trajectories view for the vehicles in the dataset
#

trajectories = (
    dataset
    .filter_labels("frames.detections", F("label") == "vehicle")
    .to_trajectories("frames.detections")
)

print(trajectories)

Parameters

field –
a frame-level label list field of any of the following types:
**kwargs – optional keyword arguments for fiftyone.core.clips.make_clips_dataset() specifying how to perform the conversion

Returns

a fiftyone.core.clips.TrajectoriesView

untag_labels(tags, label_fields=None)¶

Removes the tag from all labels in the specified label field(s) of this collection, if necessary.

Parameters

tags – a tag or iterable of tags
label_fields (None) – an optional name or iterable of names of fiftyone.core.labels.Label fields. By default, all label fields are used

untag_samples(tags)¶

Removes the tag(s) from all samples in this collection, if necessary.

Parameters: tags – a tag or iterable of tags

update_run_config(run_key, config)¶

Updates the run config for the run with the given key.

Parameters

run_key – a run key
config – a fiftyone.core.runs.RunConfig

validate_field_type(path, ftype=None, embedded_doc_type=None)¶

Validates that the collection has a field of the given type.

Parameters

path – a field name or embedded.field.name
ftype (None) – an optional field type to enforce. Must be a subclass of fiftyone.core.fields.Field
embedded_doc_type (None) – an optional embedded document type or iterable of types to enforce. Must be a subclass(es) of fiftyone.core.odm.BaseEmbeddedDocument

Raises

ValueError – if the field does not exist or does not have the expected type

validate_fields_exist(fields, include_private=False)¶

Validates that the collection has field(s) with the given name(s).

If embedded field names are provided, only the root field is checked.

Parameters

fields – a field name or iterable of field names
include_private (False) – whether to include private fields when checking for existence

Raises

ValueError – if one or more of the fields do not exist

values(field_or_expr, expr=None, missing_value=None, unwind=False, _allow_missing=False, _big_result=True, _raw=False, _field=None)¶

Extracts the values of a field from all samples in the collection.

Values aggregations are useful for efficiently extracting a slice of field or embedded field values across all samples in a collection. See the examples below for more details.

The dual function of values() is set_values(), which can be used to efficiently set a field or embedded field of all samples in a collection by providing lists of values of same structure returned by this aggregation.

Note

Unlike other aggregations, values() does not automatically unwind list fields, which ensures that the returned values match the potentially-nested structure of the documents.

You can opt-in to unwinding specific list fields using the [] syntax, or you can pass the optional unwind=True parameter to unwind all supported list fields. See Aggregating list fields for more information.

Examples:

import fiftyone as fo
import fiftyone.zoo as foz
from fiftyone import ViewField as F

dataset = fo.Dataset()
dataset.add_samples(
    [
        fo.Sample(
            filepath="/path/to/image1.png",
            numeric_field=1.0,
            numeric_list_field=[1, 2, 3],
        ),
        fo.Sample(
            filepath="/path/to/image2.png",
            numeric_field=4.0,
            numeric_list_field=[1, 2],
        ),
        fo.Sample(
            filepath="/path/to/image3.png",
            numeric_field=None,
            numeric_list_field=None,
        ),
    ]
)

#
# Get all values of a field
#

values = dataset.values("numeric_field")
print(values)  # [1.0, 4.0, None]

#
# Get all values of a list field
#

values = dataset.values("numeric_list_field")
print(values)  # [[1, 2, 3], [1, 2], None]

#
# Get all values of transformed field
#

values = dataset.values(2 * (F("numeric_field") + 1))
print(values)  # [4.0, 10.0, None]

#
# Get values from a label list field
#

dataset = foz.load_zoo_dataset("quickstart")

# list of `Detections`
detections = dataset.values("ground_truth")

# list of lists of `Detection` instances
detections = dataset.values("ground_truth.detections")

# list of lists of detection labels
labels = dataset.values("ground_truth.detections.label")

Parameters

field_or_expr –
a field name, embedded.field.name, fiftyone.core.expressions.ViewExpression, or MongoDB expression defining the field or expression to aggregate. This can also be a list or tuple of such arguments, in which case a tuple of corresponding aggregation results (each receiving the same additional keyword arguments, if any) will be returned
expr (None) –
a fiftyone.core.expressions.ViewExpression or MongoDB expression to apply to field_or_expr (which must be a field) before aggregating
missing_value (None) – a value to insert for missing or None-valued fields
unwind (False) – whether to automatically unwind all recognized list fields (True) or unwind all list fields except the top-level sample field (-1)

Returns

the list of values

write_json(json_path, rel_dir=None, include_private=False, include_frames=False, frame_labels_dir=None, pretty_print=False)¶

Writes the colllection to disk in JSON format.

Parameters

json_path – the path to write the JSON
rel_dir (None) – a relative directory to remove from the filepath of each sample, if possible. The path is converted to an absolute path (if necessary) via fiftyone.core.storage.normalize_path(). The typical use case for this argument is that your source data lives in a single directory and you wish to serialize relative, rather than absolute, paths to the data within that directory
include_private (False) – whether to include private fields
include_frames (False) – whether to include the frame labels for video samples
frame_labels_dir (None) – a directory in which to write per-sample JSON files containing the frame labels for video samples. If omitted, frame labels will be included directly in the returned JSON dict (which can be quite quite large for video datasets containing many frames). Only applicable to datasets that contain videos when include_frames is True
pretty_print (False) – whether to render the JSON in human readable format with newlines and indentations