neuralib.tracking.facemap.core.FaceMapResult

final class neuralib.tracking.facemap.core.FaceMapResult[source]

Bases: object

facemap result container

Dimension parameters:

F = number of video frames

K = number of keypoints

__init__(svd, meta, data, track_type, with_keypoints)[source]
Parameters:

  • svd (SVDVariables | None) – SVD processing outputs

  • meta (KeyPointsMeta | None) – Optional for Keypoints processing (result)

  • data (Group | None) – Optional for Keypoints processing (config)

  • track_type (Literal['keypoints', 'pupil']) – {‘keypoints’, ‘pupil’}

  • with_keypoints (bool) – if has keypoint tracking result

svd: Final[SVDVariables | None]
meta: Final[KeyPointsMeta | None]
data: Final[Group | None]
track_type: Final[Literal['keypoints', 'pupil']]
with_keypoints: Final[bool]
classmethod load(directory, track_type, *, file_pattern='')[source]

Load the facecam result from its output directory

Parameters:

  • directory (str | Path | bytes | BinaryIO | BufferedIOBase | BufferedReader) – directory contains the possible facemap output files (*.npy, *.pkl, and *.h5)

  • track_type (Literal['keypoints', 'pupil']) – {‘keypoints’, ‘pupil’}

  • file_pattern (str) – string prefix pattern to glob the facemap output file

Returns:

FaceMapResult

Return type:

Self

classmethod launch_facemap_gui(directory, with_keypoints, *, file_pattern='', env_name='neuralib')[source]

DEPRECATED. GUI view via cli.

Note that calling this method will overwrite `filenames` field in *proc.npy

Parameters:

  • directory (str | Path | bytes | BinaryIO | BufferedIOBase | BufferedReader) – directory contains the possible facemap output files (*.npy, *.pkl, and *.h5), and also the raw video file

  • with_keypoints (bool) – if has keypoint tracking result

  • file_pattern (str) – string prefix pattern to glob the facemap output file and raw avi file

  • env_name (str) – conda env name that installed the facemap package

Return type:

None

get_pupil_tracking()[source]

pupil tracking result

Return type:

PupilDict

get_pupil_area()[source]

pupil area. Array[float, F]

Return type:

ndarray

get_pupil_center_of_mass()[source]

center of mass of pupil tracking. Array[float, [F, 2]]

Return type:

ndarray

get_pupil_location_movement()[source]

Calculate the Euclidean distance from the origin for each point in a 2D array. Array[float, F]

Return type:

ndarray

eye blinking array. Array[float, F]

Return type:

ndarray

property keypoints: list[str]

list of keypoint name

get(keypoint: str) → KeyPointTrack[source]
get(keypoint: list[str]) → list[KeyPointTrack]

get a single or multiple keypoint(s)

frame_time
as_array(keypoint=None, with_outlier_filter=True, to_zscore=True, **kwargs)[source]

get keypoint(s) result as an 2D array with shape (K, F, 2). 3rd dim indicates the xy

Parameters:

  • keypoint (list[str] | str | None) – keypoint

  • with_outlier_filter (bool)

  • to_zscore (bool)

  • kwargs – pass through with_outlier_filter()

Returns:

keypoint in XY. Array[float, [K, F, 2]]

Return type:

ndarray