| Description |
% function varargout=ADAlignTrials(AlignEv,win,t0,varargin)
%
% Aligns matrices of analog data samples per trial to a given event at
% variables times during the each trial. Handy for digitally sampled
% analog data (which I refer to as AD data) stored in a repeated trial
% structure as often found in behavioral neuroscienc research. This
% doesn't use a big for loop (i.e. a for loop across trials) and so
% should be fairly quick.
%
% Example of how this is called:
% [AlX,AlY]=ADAlignTrials(Saccade,[-300 300],0,EyeX,EyeY);
%
% Where Saccade is a nTr x 1 vector of saccade onset times (in ms) for
% each trial, EyeX and EyeY are nTr x nSampsPerTrial matrices of analog
% eye positions (sampled at 1 kHz) per tial in the X and Y dimensions
% respectively. AlX and AlY will be nTr x 601 matrices of the eye
% position samples in the X and Y dimensions respectively from 300 ms
% before to 300 ms after the saccade onset in every trial, thus aligned
% on the saccade onset. t0 is the common reference time of the first
% sample in the EyeX and EyeY input matrices that Saccade is also
% appropriately referenced to.
%
% The program will align every input AD data matrix in the varargin to
% the same AlignEv each time it is run, so the number of output matrices
% will be the same as the number of input matrices in varargin.
%
% Presently this assumes a sampling rate of 1kHz for the AD data for ease
% of mental math, but allowing for different sampling rates could easily
% be done by scaling the time inputs (win, t0, AlignEv). If the aligned
% time windows go beyond the extent of the input matrices for a given
% trial, the beginning or end of the output for that trial wis padded
% with Nans.
%
% This is written not to accept NaNs in the input for AlignEv. |
