Sorting a matrix of data using another matrix of start and stop times

I have a data set of values in column with time stamps at regular intervals in the next column. I have another set of start and stop times in consecutive columns. I tried writing a function to get data in between start and stop times for it (at the bottom), but I am getting errors like input matrix must agree, inputs must have the same size etc. I have tried writing it multiple ways to get around this, but I can't. The data set is very large and there are only about ten start and stop times that I would like to get data for so it is of course not going to be equal matrices.

Any ideas?

function [behaviortimes] = datafilter(spiketimes,frequencydata)
%DATAFILTER Function returns filtered data times for frequencies given
%start and stop times of behaviors according to the frequencies
%   Start and stop times should be in two columns of a matrix
%   frequency data

if (frequencydata(:,2)>0) & (spiketimes(:,1)>0) & (spiketimes(:,2)>0);
     behaviortimes=frequencydata(frequencydata(:,2)>=(spiketimes(:,1) & frequencydata(:,2)<=(spiketimes(:,2))));
end

end

[Krispy Scripts]

Is there a way to do it without the parallel computing toolbox?

[Guillaume]

Have you looked at the answer I gave you? It does not require any toolbox.

[Krispy Scripts]

I am getting that there is an undefined function 'gte' for input arguments of type 'double'

[Guillaume]

That was a typo, it should have been ge (and le instead of lte).

Now fixed.

[Krispy Scripts]

I am getting an empty matrix for behavior times?

Answer 1

[Guillaume]

The following will tell you which row of frequencydata is in which interval of spiketimes:

 ininterval = bsxfun(@ge, frequencydata(:, 2), spiketimes(:, 1).') & bsxfun(@le, frequencydata(:, 2), spiketimes(:, 2).')

Each row of ininterval correspond to the same row of frequencydata. Each column correspond to a row of spiketimes. A 1 indicates that the row is within the interval. To reduce it to a column vector:

 isinterval = any(ininterval, 2)

And to get only the rows of frequencydata for which isinterval is true:

behaviortimes = frequencydata(isinterval, :)

As a one liner:

behaviortimes = frequencydata(any(bsxfun(@ge, frequencydata(:, 2), spiketimes(:, 1).') & bsxfun(@le, frequencydata(:, 2), spiketimes(:, 2).'), 2), :);

[Guillaume]

_"I am getting an empty matrix for behavior times?" I made another typo in the one liner, it should have been any instead of all. The detailed code was correct.

frequencydata = repmat(1:100, 2, 1)'; %demo data, data (col 1)and timestamp (col2) have same value for easy diagnostic
spiketimes = [5 10; 35 37; 78 79];
behaviortimes = frequencydata(any(bsxfun(@ge, frequencydata(:, 2), spiketimes(:, 1).') & bsxfun(@le, frequencydata(:, 2), spiketimes(:, 2).'), 2), :)  

behaviortimes = 
   5     5
   6     6
   7     7
   8     8
   9     9
  10    10
  35    35
  36    36
  37    37
  78    78
  79    79