src.canns.analyzer.spatial

Spatial analysis utilities for neural activity data.

This module provides functions for analyzing spatial patterns in neural data, particularly for computing firing fields and spatial smoothing operations.

Functions

compute_firing_field(A, positions, width, height, M, K)

Compute spatial firing fields for neural population activity.

gaussian_smooth_heatmaps(heatmaps[, sigma])

Apply Gaussian smoothing to spatial heatmaps without mixing channels.

Module Contents

src.canns.analyzer.spatial.compute_firing_field(A, positions, width, height, M, K)[source]

Compute spatial firing fields for neural population activity.

This function bins neural activity into a 2D spatial grid based on the animal’s position, creating a heatmap for each neuron showing where it fires most strongly. Uses Numba JIT compilation for high performance.

Parameters:

  • A (np.ndarray) – Neural activity array of shape (T, N) where T is the number of time steps and N is the number of neurons.

  • positions (np.ndarray) – Position data of shape (T, 2) containing (x, y) coordinates at each time step.

  • width (float) – Width of the spatial environment.

  • height (float) – Height of the spatial environment.

  • M (int) – Number of bins along the width dimension.

  • K (int) – Number of bins along the height dimension.

Returns:

Heatmaps array of shape (N, M, K) containing the average

firing rate of each neuron in each spatial bin.

Return type:

np.ndarray

Example

>>> activity = np.random.rand(1000, 30)  # 1000 timesteps, 30 neurons
>>> positions = np.random.rand(1000, 2) * 5.0  # Random walk in 5x5 space
>>> heatmaps = compute_firing_field(activity, positions, 5.0, 5.0, 50, 50)
>>> heatmaps.shape
(30, 50, 50)
src.canns.analyzer.spatial.gaussian_smooth_heatmaps(heatmaps, sigma=1.0)[source]

Apply Gaussian smoothing to spatial heatmaps without mixing channels.

This function applies Gaussian filtering to each heatmap independently, preserving zero values (unvisited spatial bins) and only smoothing regions with activity.

Parameters:

  • heatmaps (np.ndarray) – Array of shape (N, M, K) where N is the number of neurons/channels and (M, K) is the spatial grid size.

  • sigma (float, optional) – Standard deviation for Gaussian kernel. Defaults to 1.0.

Returns:

Smoothed heatmaps with the same shape as input. Zero values

in the original heatmaps are preserved.

Return type:

np.ndarray

Example

>>> heatmaps = np.random.rand(30, 50, 50)
>>> smoothed = gaussian_smooth_heatmaps(heatmaps, sigma=1.5)
>>> smoothed.shape
(30, 50, 50)