On Dec 24, 1:24 am, "Matt J " <mattjacREM...@THISieee.spam> wrote: > Aseman <andali...@gmail.com> wrote in message <71c2925e-ad3a-430f-b9b3-18453386c...@t8g2000yqg.googlegroups.com>... > > hi > > > I should apply my own function fun to the columns of a matrix > > A(200,000 by 1000). How to do this without cellfun or for loops? > > any help will be greatly appreciated. > > =============== > > What's the function?
Dear Matt This is actually the problem my colleague, Rahill, seek advice about it. You kindly answered to his question in her past post. Your code works more faster on each column of the matrix consisting of 180,000 elements, but how we could run it over all 744 columns without any for loop?
Here is the original problem and your answer.
___________________________________________________________ hello everybody I am working with a 180,000 by 744 data matrix. At this time, I should work with each column of this matrix. I should divide each column to 16 segments overlapping by 75%. To reduce the complexity, I have tried to avoid any loop in my code, as the following
however, I do not know how to apply the above code to all of the 744 columns without using any loop. I tried to do so, by concatenating the columns, but in this way, MATLAB crashes by 'out of memory' prompt.
any help will be appreciated Rahill
Matt One possibility: A= mod( bsxfun(@plus,(1:12).',3*(0:4))-1,15)+1 __________________________________________________________