compute_features

Compute track features.

This module contains functions computing different features which will be used as input of the deep-learning network (for both training and testing).

compute_features.compute_all_features(track_df)

Computes all features necessary for the inputs of the deep-learning network. The features are directly added to the dataframe of trajectories.

Parameters:

track_df (pd.DataFrame) – All trajectories as a dataframe with keys: x, y, frame, data_folder, track_id.

compute_features.compute_angles(displ_x, displ_y)

Computes angles using two consecutive displacements using x and y 1D-displacements.

Parameters:

  • displ_x (np.array) – x displacements of a given track.

  • displ_y (np.array) – y displacements of a given track.

Returns:

Angles computed along the track.

Return type:

np.array

compute_features.compute_displacements(track, delta)

Computes x and y displacements between each pair of points i and i+delta.

Parameters:

  • track (pd.DataFrame) – Given track where displacements will be computed along its entire length.

  • delta (int) – Lag interval between frames.

Returns:

Computed (x displacements, y displacements) along the track.

Return type:

(np.array, np.array)

compute_features.compute_dist(displ_x, displ_y)

Computes distances based on x and y displacements.

Parameters:

  • displ_x (np.array) – x displacements of a given track.

  • displ_y (np.array) – y displacements of a given track.

Returns:

Computed distances along the track.

Return type:

np.array

compute_features.compute_mean_distances(track, delta, num=1)

Computes a mean of distances from point i-num to point i+num.

Parameters:

  • track (pd.DataFrame) – Given track where mean of distances will be computed along its entire length.

  • delta (int) – Lag interval between frames while measuring the displacements.

  • num (int) – Number of neighboring point for the average. [Defaults: +/- 1 close neighbors]

Returns:

The mean of distances.

Return type:

np.array