stnmf.preprocessing.spiketriggered.stimulusensemble

stnmf.preprocessing.spiketriggered.stimulusensemble(stimulus, spikes, sta_temp, continuous=False, min_spikes_per_bin=1, dtype='float32')

Collect the effective spike-triggered stimulus ensemble (STE)

Parameters:

  • stimulus (iterator or array_like) – This iterator should yield a (x,y,trial_length) array_like for each trial, where trial_length matches the number of elements of spikes. The stimulus should be cropped around the cell’s receptive field. See function window

  • spikes ((num_trials,trial_length,num_cells) array_like) – Binned spike counts of one cell, i.e. num_cells has to be 1 (see notes). If two dimensional, assume num_cells == 1

  • sta_temp ((tau,) array_like) – Temporal profile of the spike-triggered average. See function profiles

  • continuous (bool, optional) – Specify if trials should be treated as connected chunks of a continuous recording or if they are separated. Default is False

  • min_spikes_per_bin (int, optional) – Specify the minimum number of spikes per bin to be considered. This is useful for cells with high baseline firing rate. Default is 1

  • dtype (str or numpy.dtype, optional) – Data type in which to perform and return the computations. Default is ‘float32’

Returns:

ste ((x,y,s) numpy.ndarray) – Effective Spike-triggered stimulus ensemble, where s is the number of spikes

Raises:

NotImplementedError : – If the last dimension of the spikes array is greater than one

Notes

The current implementation only supports spikes from one cell at a time.

Each spike yields one spike-triggered stimulus, that is, multiple spikes in one bin result in identical successive spike-triggered stimulus frames in the STE.