I have to correlate one time-serie with thousands of others. So far im using a loop which looks approximately like this:
r(i) = corr(my_timeserie,other_timeserie(i))
The problem is that it takes forever to run, i was wondering if there was a mass-univariate way to do this without a loop. I thought about using xcorr , but the results is way too big and often create a memory crash.