Matrix Capacity

For requesting general information about EViews, sharing your own tips and tricks, and information on EViews training or guides.

Moderators: EViews Gareth, EViews Moderator

CharlieEVIEWS
Posts: 202
Joined: Tue Jul 17, 2012 9:47 am

Matrix Capacity

Postby CharlieEVIEWS » Tue Jul 21, 2015 8:00 am

Dear all,

Could somebody possibly provide some details on the maximum permissible number of matrix elements? Cant see any information on http://www.eviews.com/EViews9/ev9capacity.html? It seems a 625m element matrix is too large. I guess I could write a short program to sequentially test the boundary with increasing increments, but might be not be the fastest thing to do :)

Is it a a hard threshold as opposed to one that's curtailed by hardware/subject to system requirements?

Best wishes as always!

Charlie

EViews Gareth
Fe ddaethom, fe welon, fe amcangyfrifon
Posts: 13600
Joined: Tue Sep 16, 2008 5:38 pm

Re: Matrix Capacity

Postby EViews Gareth » Tue Jul 21, 2015 8:13 am

I believe rows*(cols+1)<268,435,424

CharlieEVIEWS
Posts: 202
Joined: Tue Jul 17, 2012 9:47 am

Re: Matrix Capacity

Postby CharlieEVIEWS » Tue Jul 21, 2015 8:41 am

Interesting crash bug: when i just tried to create a second (large) matrix to test near to the limit, i got a 'stopped working' message. Undoubtedly due to the fact I'm on a 4gb laptop only, but a warning message might be cooler than a crash bug :P

As a test, I then ran:

Code: Select all

wfcreate u 1 for !a = 1 to 999999 matrix(10000,10000) matrix_!a next
And I got to 3 :oops:

EViews Gareth
Fe ddaethom, fe welon, fe amcangyfrifon
Posts: 13600
Joined: Tue Sep 16, 2008 5:38 pm

Re: Matrix Capacity

Postby EViews Gareth » Tue Jul 21, 2015 8:51 am

It is very difficult to error politely when you run out of memory :(

startz
Non-normality and collinearity are NOT problems!
Posts: 3797
Joined: Wed Sep 17, 2008 2:25 pm

Re: Matrix Capacity

Postby startz » Tue Jul 21, 2015 8:57 am

Interestingly, I did not get a crash. After 80 matrices things slowed down a lot. That's about right on my 64GB machine. Wonder if the bug has to do with 32-bit versus 64-bit.

CharlieEVIEWS
Posts: 202
Joined: Tue Jul 17, 2012 9:47 am

Re: Matrix Capacity

Postby CharlieEVIEWS » Tue Jul 21, 2015 9:35 am

I'm running 64bit EE on this system, which is just curtailed by RAM size. I bet if you changed 10000 to ~16000 and let it run for a few hours, you'd see the same crash even with 64gb of ram...

startz
Non-normality and collinearity are NOT problems!
Posts: 3797
Joined: Wed Sep 17, 2008 2:25 pm

Re: Matrix Capacity

Postby startz » Tue Jul 21, 2015 9:44 am

So it sounds like a bug from a failed memory allocation. My machine would just take longer to hit.

Perhaps for debugging purposes they should dig out the Apple II that the program ran on originally. I think it had 48k of memory.


Return to “General Information and Tips and Tricks”

Who is online

Users browsing this forum: No registered users and 2 guests