Compiling matlab code using the optimization toolbox

My question relates to using the Matlab C Compiler to compile and run programs utilizing functions from Matlab's Optimization Toolbox. Specifically I am using the lsqnonlin() function. My actual program does very little reading in or writing out if I run it in Matlab; however, if I compile it and run the C code (on a cluster where I can run in parallel), then there is a heavy amount of input/output which is slowing down the network it's running on. My thought is that perhaps the compiled version of lsqnonlin() necessitates input/output during individual optimization cycles/function-evaluations, whereas the uncompiled matlab version only writes out to a file after lsqnonlin() is done evaluating. Does anyone know if this is the case or have any alternative thoughts about what might be happening? Thanks!

Answers (0)

Categories

Asked:

on 26 Nov 2012

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!