I mean indexing into tmp for assignment like:

tmp(ind1,ind2) = whatever

–Loren

That’s simple. YOu are asking for the allocation by saying “tmp=”

Memory will always be allocated for tmp under that sort of situation. You need to index into the tmp variable you preallocated for it to reuse the memory, even though tmp is shared from the parent to the nested function.

–Loren

function fh= get_fctr(X);
n=size(X,1);
tmp=zeros(n,1);
fh=@fcn_implem;
function [res1,res2]=function_implem(i)
 tmp=known_function0(X,i);%here!this is where allocation is %happening and I don't understand why!
 res1=known_function1(tmp,i);
 res2=known_function2(tmp,i);
end
end
%now in a 'main' function I am calling this functor several %times.
afunc=get_fctr(anX);
for i=1:largenumber
store_results(:,i)=afunc(i);
end

and for some reason I don’t know why it needs to reallocate the tmp inside the function every time instead of just once when I do the “afunc=get_fctr(anX)” call.
Hope this is clear enough? Thanks!

Yes, object properties also take advantage of the copy-on-write behavior. If you are on windows, you can watch the task manager memory plot as you assign to an object’s property and you should see no memory spike.

–Loren

I have a specific question:
Does Matlab implement copy-on-write even for arrays passed in as arguments when constructing an object? From profiling output I’ve observed, I am lead to suspect that it there may be some kind of copy-on-access behavior for large arrays passed in to the constructor of class.

Sorry for the cross-post, but I didn’t get any responses when I posted a more detailed version of this question on the newsgroup at:

http://www.mathworks.com/matlabcentral/newsreader/view_thread/284793#755110

(I have a code snippet in that post along with profiling output).

Thanks for reading,

Arvind

I recommend you look at tech support guide to memory management. It has programming tips, ways to set up your computer environment etc.

–Loren

I am running some mex file and getting out of memory error.
When i use “memory” command from matlab prompt i can see matlab is using only around 10% of the total physical memory available. Also, when i add up all the arrays that i am declaring inside mex file, it becomes only few percent of total memory available.

Now my question is – can we ask matlab to use more memory from the available RAM or is there any other alternate without buying a new computer/etc.

Thanks
Anis

Not sure what you are asking. Sounds like you’ve figured out how to get better performance.

–Loren