Simbiology memory hog
in [Matlab]
|
Prev: OPEN AND CLOSE THE APPLICATION EXECUTABLE, an error in MS-DOSand it is MCR OF MATLAB
Next: Matlab assistance on variable names
From: Henry Lester on 21 May 2010 19:05 I'm **not** an accomplished Matlab programmer but have written a Simbiology model with ~50 compartments and ~19 parameters. I'm trying to use simulated annealing. Matlab seems to be adding ~ 1 MB per iteration to its memory use, severely limiting my ability to use simulated annealing. I'm assuming it's a simple error but have tried various "clear" and "delete" statements to no avail. Any ideas? Where do I send the files for further inspection?
From: Steven Lord on 21 May 2010 21:06 "Henry Lester" <lester(a)caltech.edu> wrote in message news:ht73jl$89a$1(a)fred.mathworks.com... > I'm **not** an accomplished Matlab programmer but have written a > Simbiology model with ~50 compartments and ~19 parameters. > > I'm trying to use simulated annealing. > > Matlab seems to be adding ~ 1 MB per iteration to its memory use, severely > limiting my ability to use simulated annealing. > > I'm assuming it's a simple error but have tried various "clear" and > "delete" statements to no avail. > > Any ideas? Where do I send the files for further inspection? Contact Technical Support. See my signature for instructions on how to do so. -- Steve Lord slord(a)mathworks.com comp.soft-sys.matlab (CSSM) FAQ: http://matlabwiki.mathworks.com/MATLAB_FAQ To contact Technical Support use the Contact Us link on http://www.mathworks.com
From: Arthur Goldsipe on 22 May 2010 00:17
"Henry Lester" <lester(a)caltech.edu> wrote in message <ht73jl$89a$1(a)fred.mathworks.com>... > I'm **not** an accomplished Matlab programmer but have written a Simbiology model with ~50 compartments and ~19 parameters. > > I'm trying to use simulated annealing. > > Matlab seems to be adding ~ 1 MB per iteration to its memory use, severely limiting my ability to use simulated annealing. > > I'm assuming it's a simple error but have tried various "clear" and "delete" statements to no avail. > > Any ideas? Where do I send the files for further inspection? Hi Henry, I'm a developer for SimBiology. I'm very interested in seeing what you're doing. Feel free to send your files to me directly as well. In the meantime, I may know what the problem is. Are you calling sbiomodel or sbmlimport in every iteration? To prevent accidental loss of data, SimBiology models have the unusual behavior of remainng in memory until they are explicity deleted. Even if you call clear or overwrite a variable, the model will be stored on the SimBiology root object. This object can be accessed via the sbioroot function. To fix this problem, call delete on a variable containing a model before assigning a new value to it. Or, if you wish to delete ALL models in memory, you can also call the sbioreset function. Hope that helps! -Arthur |