gridcells.analysis.fields - grid field related analysis

The fields module contains routines to analyse spiking data either from experiments involoving a rodent running in an arena or simulations involving an animat running in a simulated arena.

Functions

gridnessScore(rateMap, arenaDiam, h, ...)	Calculate gridness score of a spatial firing rate map.
occupancy_prob_dist(arena, pos)	Calculate a probability distribution for animal positions in an arena.
spatialAutoCorrelation(rateMap, arenaDiam, h)	Compute autocorrelation function of the spatial firing rate map.
spatialRateMap(spikeTimes, positions, arena, ...)	Compute spatial rate map for spikes of a given neuron.

gridcells.analysis.fields.gridnessScore(rateMap, arenaDiam, h, corr_cutRmin)

Calculate gridness score of a spatial firing rate map.

Parameters:

rateMap : np.ndarray

Spatial firing rate map.

arenaDiam : float

The diameter of the arena.

h : float

Precision of the spatial firing rate map.

Returns:

G : float

Gridness score.

crossCorr : np.ndarray

An array containing cross correlation values of the rotated autocorrelations, with the original autocorrelation.

angles : np.ndarray

An array of angles corresponding to the crossCorr array.

Notes

This function computes gridness score accoring to [R1]. The auto correlation of the firing rate map is rotated in 3 degree steps. The resulting gridness score is the difference between a minimum of cross correlations at 60 and 90 degrees, and a maximum of cross correlations at 30, 90 and 150 degrees.

The center of the auto correlation map (given by corr_cutRmin) is removed from the map.

References

[R1]	(1, 2) Hafting, T. et al., 2005. Microstructure of a spatial map in the entorhinal cortex. Nature, 436(7052), pp.801-806.

gridcells.analysis.fields.occupancy_prob_dist(arena, pos)

Calculate a probability distribution for animal positions in an arena.

Parameters:

arena : Arena

Arena the animal was running in.

pos : Position2D

Positions of the animal.

Returns:

dist : numpy.ndarray

Probability distribution for the positional data, given the discretisation of the arena. The first dimension is the y axis, the second dimension is the x axis. The shape of the distribution is equal to the number of items in the discretised edges of the arena.

gridcells.analysis.fields.spatialAutoCorrelation(rateMap, arenaDiam, h)

Compute autocorrelation function of the spatial firing rate map.

This function assumes that the arena is a circle and masks all values of the autocorrelation that are outside the arenaDiam.

Warning

This function will undergo serious interface changes in the future.

Parameters:

rateMap : np.ndarray

Spatial firing rate map (2D). The shape should be (arenadiam/h+1, arenadiam/2+1).

arenaDiam : float

Diameter of the arena.

h : float

Precision of the spatial firing rate map.

Returns:

corr : np.ndarray

The autocorrelation function, of shape (arenadiam/h*2+1, arenaDiam/h*2+1)

xedges, yedges : np.ndarray

Values of the spatial lags for the correlation function. The same shape as corr.shape[0].

gridcells.analysis.fields.spatialRateMap(spikeTimes, positions, arena, sigma)

Compute spatial rate map for spikes of a given neuron.

Preprocess neuron spike times into a smoothed spatial rate map, given arena parameters. Both spike times and positional data must be aligned in time! The rate map will be smoothed by a gaussian kernel.

Parameters:

spikeTimes : np.ndarray

Spike times for a given neuron.

positions : gridcells.core.Position2D

Positional data for these spikes. The timing must be aligned with spikeTimes

arena : gridcells.core.Arena

The specification of the arena in which movement was carried out.

sigma : float

Standard deviation of the Gaussian smoothing kernel.

Returns:

rateMap : np.ma.MaskedArray

The 2D spatial firing rate map. The shape will be determined by the arena type.