[galib] need help with the examples in galib for master slave GA using PVM

[ratnoo saroj]

Dear all

Hi, I am a research student in Jawaharlal Nehru
University in India. I am working on parallel GA
models in data mining. I have implemented a master
slave GA. My program runs well if I keep a small
population size like 20.  In case of a popluation size
about 50 and more, the program works OK up to few
generations like (1000) and then it slows down
enormously. I have debugged my program nothing seems
wrong.

Finally, I ran the master slave implementation given
in galib245/examples/pvmind having a population of 150
and generations 2000. Even this example behaves the
same way. It runs ok upto some generation and then
slows down a lot. I have run the same example earlier
and it worked fine on an older version of LINUX 

I am working in galib245 and Red Hat Enterprise WS
release 3. While booting up, Linux shows that it needs
a RAM of 256 MB and only 240 is detected and then
normal start up begins.

Can anyone give any clue about this weird behaviour.
Can it be the the RAM size?

Please help.

Saroj


Send instant messages to your online friends http://uk.messenger.yahoo.com