fiftyone.core.stages.SortBySimilarity

class documentation

class SortBySimilarity(ViewStage): (source)

Constructor: SortBySimilarity(query, k, reverse, dist_field, ...)

Sorts a 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

stage = fo.SortBySimilarity(query_id, k=5)
view = dataset.add_stage(stage)

#
# 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)

stage = fo.SortBySimilarity(query, k=5)
view = dataset.add_stage(stage)

#
# Sort samples by their similarity to a text prompt
#

query = "kites high in the air"

stage = fo.SortBySimilarity(query, k=5)
view = dataset.add_stage(stage)

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	the number of matches to return. By default, the entire collection is sorted
reverse	whether to sort by least similarity (True) or greatest similarity (False). Some backends may not support least similarity
dist_field	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	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

Method	`__init__`	Undocumented
Method	`to_mongo`	Returns the MongoDB aggregation pipeline for the stage.
Method	`validate`	Validates that the stage can be applied to the given collection.
Property	`brain_key`	The brain key of the similarity index to use.
Property	`dist_field`	The field to store similarity distances, if any.
Property	`k`	The number of matches to return.
Property	`query`	The query.
Property	`reverse`	Whether to sort by least similarity.
Class Method	`_params`	Returns a list of JSON dicts describing the stage's supported parameters.
Method	`_kwargs`	Returns a list of `[name, value]` lists describing the parameters of this stage instance.
Method	`_make_pipeline`	Undocumented
Instance Variable	`_brain_key`	Undocumented
Instance Variable	`_dist_field`	Undocumented
Instance Variable	`_is_prompt`	Undocumented
Instance Variable	`_k`	Undocumented
Instance Variable	`_pipeline`	Undocumented
Instance Variable	`_query`	Undocumented
Instance Variable	`_query_kwarg`	Undocumented
Instance Variable	`_reverse`	Undocumented
Instance Variable	`_state`	Undocumented

Inherited from ViewStage:

Method	`__eq__`	Undocumented
Method	`__repr__`	Undocumented
Method	`__str__`	Undocumented
Method	`get_edited_fields`	Returns a list of names of fields or embedded fields that may have been edited by the stage, if any.
Method	`get_excluded_fields`	Returns a list of fields that have been excluded by the stage, if any.
Method	`get_filtered_fields`	Returns a list of names of fields or embedded fields that contain arrays have been filtered by the stage, if any.
Method	`get_group_expr`	Returns the dynamic group expression for the given stage, if any.
Method	`get_media_type`	Returns the media type outputted by this stage when applied to the given collection, if and only if it is different from the input type.
Method	`get_selected_fields`	Returns a list of fields that have been selected by the stage, if any.
Method	`load_view`	Loads the `fiftyone.core.view.DatasetView` containing the output of the stage.
Property	`has_view`	Whether this stage's output view should be loaded via `load_view` rather than appending stages to an aggregation pipeline via `to_mongo`.
Property	`outputs_dynamic_groups`	Whether this stage outputs or flattens dynamic groups.
Class Method	`_from_dict`	Creates a `ViewStage` instance from a serialized JSON dict representation of it.
Method	`_needs_frames`	Whether the stage requires frame labels of video samples to be attached.
Method	`_needs_group_slices`	Whether the stage requires group slice(s) to be attached.
Method	`_serialize`	Returns a JSON dict representation of the `ViewStage`.
Instance Variable	`_uuid`	Undocumented

def __init__(self, query, k=None, reverse=False, dist_field=None, brain_key=None, _state=None): (source) ¶

Undocumented

def to_mongo(self, _): (source) ¶

overrides fiftyone.core.stages.ViewStage.to_mongo

Returns the MongoDB aggregation pipeline for the stage.

Only usable if has_view is False.

Parameters
_	Undocumented
sample_collection	the `fiftyone.core.collections.SampleCollection` to which the stage is being applied
Returns
a MongoDB aggregation pipeline (list of dicts)