placecell.neural

placecell.neural#

Neural data loading and deconvolution.

Functions

build_event_index_dataframe(unit_ids, S_list)

Build event index DataFrame from spike trains.

load_calcium_traces(neural_path[, trace_name])

Load traces from a Minian-style zarr store as a DataArray.

run_deconvolution(C_da, unit_ids, g, ...[, ...])

Run OASIS deconvolution on calcium traces.
placecell.neural.load_calcium_traces(neural_path: Path, trace_name: str = 'C') → DataArray#

Load traces from a Minian-style zarr store as a DataArray.

Parameters:

  • neural_path (Path) – Directory containing <trace_name>.zarr.

  • trace_name (str) – Base name of the zarr group (e.g. "C" or "C_lp"). Also used as the variable name if the zarr contains a Dataset.

Returns:

DataArray with dimensions (‘unit_id’, ‘frame’).

Return type:

xr.DataArray

placecell.neural.run_deconvolution(C_da: Any, unit_ids: list[int], g: tuple[float, float], baseline: float | str, penalty: float, s_min: float, progress_bar: Any = None) → tuple[list[int], list[ndarray], list[ndarray]]#

Run OASIS deconvolution on calcium traces.

Parameters:

  • C_da (xarray.DataArray) – Calcium traces with dimensions (unit_id, frame).

  • unit_ids (list[int]) – List of unit IDs to process.

  • g (tuple[float, float]) – AR(2) coefficients for OASIS.

  • baseline (float or str) – Baseline correction. Use ‘pXX’ for percentile (e.g., ‘p10’) or numeric value.

  • penalty (float) – Sparsity penalty for OASIS.

  • s_min (float) – Minimum event size threshold.

  • progress_bar (optional) – tqdm progress bar wrapper (e.g., tqdm.notebook.tqdm).

Returns:

  • good_unit_ids (list[int]) – Unit IDs that were successfully deconvolved.

  • S_list (list[np.ndarray]) – Spike trains.

Return type:

tuple[list[int], list[ndarray], list[ndarray]]

placecell.neural.build_event_index_dataframe(unit_ids: list[int], S_list: list[ndarray]) → DataFrame#

Build event index DataFrame from spike trains.

Parameters:

  • unit_ids (list[int]) – Unit IDs corresponding to each spike train.

  • S_list (list[np.ndarray]) – List of spike train arrays.

Returns:

Event index with columns: unit_id, frame, s.

Return type:

pd.DataFrame

On this page