Parallel Computing GA algorithm freezes after several generations

Hello,

I've been running into a problem using the GA algorithm in parallel that it freeze sometime into the iteration after having successfully passed through several (around 10-15) generations.

The command window displays this status:

In globaloptim\private\createAnonymousFcn>@(x)fcn(x,FcnArgs{:}) at 11
In globaloptim\private\fcnvectorizer>(parfor body) at 17
In parallel_function>make_general_channel/channel_general at 914
In remoteParallelFunction at 38

It will remain in this function forever (I haven't waited that long yet, but seemingly so at least). I'm not sure what would cause the ga to fail at this point late in the run after several generations, any advice would be appreciated. Thanks!

[Shruti Sapre]

Hi Brandon,

I understand that the GA algorithm you run stops after going through several generations. Could you provide some more information on the data being used?

In order to help debug this, you could also try turning the ‘Display’ option to ‘diagnose’ while running the algorithm. This will display the following fields and will hopefully provide some information about the issue:

1. Generation — Generation number

2. f-count — Cumulative number of fitness function evaluations

3. Best f(x) — Best fitness function value

4. Mean f(x) — Mean fitness function value

-Shruti