123456789101112131415161718192021 |
- function [data, peakresp] = normalize(data,trials,t,peakwin)
- % this function normalizes data by first smoothing the psth of the
- % trials indicated by the trials array (ONLY INCLUDE TRIALS THAT HAVE A
- % PEAK RESPONSE, i.e. trials in which there was a stimulus). it looks for
- % the peak in this smoothed psth within a given peak window, and divides
- % all data by the size of this peak
- %
- % input:
- % - data (time by trials, should be already BASELINE CORRECTED)
- % - trials (array of booleans indicating which trials to include in the
- % psth for peak calculation)
- % - time axis in same unit as peak window
- % - [start end] peak window
- %
- % Rob Teeuwen 20190815
- psth = nanmean(data(:,trials==1),2);
- psths = smooth(psth,20,'lowess');
- peakresp = max(psths(t>peakwin(1) & t<peakwin(2)));
- data = data./peakresp;
|