replay_trajectory_classification.likelihoods.multiunit_likelihood_integer.estimate_multiunit_likelihood_integer#

estimate_multiunit_likelihood_integer(multiunits: ndarray, encoding_marks: ndarray, mark_std: ndarray, place_bin_centers: ndarray, encoding_positions: ndarray, position_std: ndarray, occupancy: ndarray, mean_rates: ndarray, summed_ground_process_intensity: ndarray, bin_diffusion_distances: ndarray, edges: list[ndarray], max_mark_diff: int = 6000, set_diag_zero: bool = False, is_track_interior: ndarray | None = None, time_bin_size: int = 1, block_size: int = 100, ignore_no_spike: bool = False, disable_progress_bar: bool = False, use_diffusion: bool = False) → ndarray[source]#

Estimates the likelihood of position bins given multiunit marks.

Parameters:

  • multiunits (np.ndarray, shape (n_decoding_time, n_marks, n_electrodes))

  • encoding_marks (np.ndarray, shape (n_encoding_spikes, n_marks, n_electrodes))

  • mark_std (float) – Amount of smoothing for mark features. Standard deviation of kernel.

  • place_bin_centers (np.ndarray, shape (n_bins, n_position_dims))

  • encoding_positions (np.ndarray, shape (n_encoding_spikes, n_position_dims))

  • position_std (float or array_like, shape (n_position_dims,)) – Amount of smoothing for position. Standard deviation of kernel.

  • occupancy (np.ndarray, (n_bins,))

  • mean_rates (list, len (n_electrodes,))

  • summed_ground_process_intensity (np.ndarray, shape (n_bins,))

  • bin_diffusion_distances (np.ndarray, shape (n_bins, n_bins))

  • edges (list of np.ndarray)

  • max_mark_diff (int) – Maximum difference between mark features

  • set_diag_zero (bool) – Remove influence of the same mark in encoding and decoding.

  • is_track_interior (None or np.ndarray, shape (n_bins_x, n_bins_y))

  • time_bin_size (float) – Size of time steps

  • block_size (int) – Size of data to process in chunks

  • ignore_no_spike (bool) – Set contribution of no spikes to zero

  • disable_progress_bar (bool) – If False, a progress bar will be displayed.

  • use_diffusion (bool) – Respect track geometry by using diffusion distances

Returns:

log_likelihood

Return type:

np.ndarray, shape (n_time, n_bins)