allensdk.brain_observatory.behavior.data_objects.stimuli.stimulus_templates module

class allensdk.brain_observatory.behavior.data_objects.stimuli.stimulus_templates.StimulusImage(warped: ndarray, unwarped: ndarray, name: str)

Bases: object

Container class for image stimuli

property name

class allensdk.brain_observatory.behavior.data_objects.stimuli.stimulus_templates.StimulusImageFactory

Bases: object

Factory for StimulusImage

static from_processed(warped: ndarray, unwarped: ndarray, name: str) → StimulusImage

Creates a StimulusImage from processed input (usually nwb). Image has already been warped and preprocessed

from_unprocessed(input_array: ndarray, name: str) → StimulusImage

Creates a StimulusImage from unprocessed input (usually pkl). Image needs to be warped and preprocessed

class allensdk.brain_observatory.behavior.data_objects.stimuli.stimulus_templates.StimulusMovieFrameFactory

Bases: StimulusImageFactory

class allensdk.brain_observatory.behavior.data_objects.stimuli.stimulus_templates.StimulusMovieTemplateFactory

Bases: StimulusTemplateFactory

Modified stimulus template factory to read a natural movie.

static from_unprocessed(movie_name: str, movie_frames: List[ndarray], n_workers: int | None = None) → StimulusTemplate

Create StimulusTemplate from pkl or unprocessed input. Stimulus templates created this way need to be processed to acquire unwarped versions of the movie frames presented.

NOTE: Warped movie frames display what was seen on a monitor by a subject. Unwarped images display a ‘diagnostic’ version of the stimuli to be presented.

Parameters:

movie_namestr

The name of the image set. Example:

Natural_Images_Lum_Matched_set_TRAINING_2017.07.14

movie_framesList[np.ndarray]

A list of image arrays

n_workersint, Optional

Number of processes to spawn to warp movie frames. Defaults to os.cpu_count if not specified.

Returns:

StimulusTemplate

A StimulusTemplate object

class allensdk.brain_observatory.behavior.data_objects.stimuli.stimulus_templates.StimulusTemplate(image_set_name: str, images: List[StimulusImage])

Bases: object

Container class for a collection of image stimuli

property image_names: List[str]

property image_set_name: str

property images: List[StimulusImage]

items()

keys()

to_dataframe(index_name: str = 'image_name', index_type: str = 'str') → DataFrame

Convert the collection of stimulus templates to a dataframe.

Parameters:

index_namestr

Name apply to the index column of the data frame to return. Defaults to “image_name”.

index_typestr

Type of the index of the dataframe to set. Defaults to ‘str’.

Returns:

dfpandas.DataFrame

Dataframe containing an index and two columns: warped, unwarped. (un)warped columns return as a list of lists representing the displayed image/frame.

values()

class allensdk.brain_observatory.behavior.data_objects.stimuli.stimulus_templates.StimulusTemplateFactory

Bases: object

Factory for StimulusTemplate

static from_processed(image_set_name: str, image_attributes: List[dict], unwarped: List[ndarray], warped: List[ndarray]) → StimulusTemplate

Create StimulusTemplate from nwb or other processed input. Stimulus templates created this way DO NOT need to be processed to acquire unwarped versions of the images presented.

NOTE: The ordering of image_attributes, unwarped, and warped matter!

NOTE: Warped images display what was seen on a monitor by a subject. Unwarped images display a ‘diagnostic’ version of the stimuli to be presented.

Parameters:

image_set_namestr

The name of the image set. Example:

Natural_Images_Lum_Matched_set_TRAINING_2017.07.14

image_attributesList[dict]

A list of dictionaries containing image metadata. Must at least contain the key:

image_name

But will usually also contain:

image_category, orientation, phase, spatial_frequency, image_index

unwarpedList[np.ndarray]

A list of unwarped image arrays

warpedList[np.ndarray]

A list of warped image arrays

Returns:

StimulusTemplate

A StimulusTemplate object

static from_unprocessed(image_set_name: str, image_attributes: List[dict], images: List[ndarray]) → StimulusTemplate

Create StimulusTemplate from pkl or unprocessed input. Stimulus templates created this way need to be processed to acquire unwarped versions of the images presented.

NOTE: The ordering of image_attributes and images matter!

NOTE: Warped images display what was seen on a monitor by a subject. Unwarped images display a ‘diagnostic’ version of the stimuli to be presented.

Parameters:

image_set_namestr

The name of the image set. Example:

Natural_Images_Lum_Matched_set_TRAINING_2017.07.14

image_attributesList[dict]

A list of dictionaries containing image metadata. Must at least contain the key:

image_name

But will usually also contain:

image_category, orientation, phase, spatial_frequency, image_index

imagesList[np.ndarray]

A list of image arrays

Returns:

StimulusTemplate

A StimulusTemplate object