replay_trajectory_classification.likelihoods.multiunit_likelihood.fit_multiunit_likelihood#
- fit_multiunit_likelihood(position: ndarray, multiunits: ndarray, place_bin_centers: ndarray, mark_std: ndarray | float, position_std: ndarray | float, is_track_boundary: ndarray | None = None, is_track_interior: ndarray | None = None, edges: list[ndarray] | None = None, block_size: int = 100, use_diffusion: bool = False, **kwargs) dict[source]#
Fits the clusterless place field model.
- Parameters:
position (np.ndarray, shape (n_time, n_position_dims))
multiunits (np.ndarray, shape (n_time, n_marks, n_electrodes))
place_bin_centers (np.ndarray, shape (n_bins, n_position_dims))
mark_std (float or array_like, shape (n_marks,)) – Amount of smoothing for the mark features. Standard deviation of kernel.
position_std (float or array_like, shape (n_position_dims,)) – Amount of smoothing for position. Standard deviation of kernel.
is_track_boundary (None or np.ndarray, shape (n_bins,))
is_track_interior (None or np.ndarray, shape (n_bins,))
edges (None or list of np.ndarray)
block_size (int) – Size of data to process in chunks
use_diffusion (bool) – Use diffusion to respect the track geometry.
- Returns:
encoding_model
- Return type: