I'm running a loop that create feature classes at a geodatabase. After creating 221 datasets I get the error message "pyscripter out of memory". Is there a way to clear the physical memory of computer in order to resume with the script ? I've alredy tried :
gc.enable
clear = lambda: os.system('cls')
clear()
gc.collect()
I would start first looking for a place to figure out at which time the memory gets full. The memory consumption can be monitored with an extra module psutil, which is a great cross-platform library for obtaining information on running processes and system utilization (i.e., CPU, memory, disks, network). I guess you could also use any other logging techniques in your code that Python has to control the memory use.
Another general thing you might want to try is to use a module subprocess; there is a great post on how to use this module here. You basically would try to spawn new processes, then connect to their pipes, and obtain their return codes.
If you're creating an artifact in memory, like a FeatureLayer, with each iteration, you could be piling up the memory you use. I would look for places where arcpy.Delete_management() could be used, rather than the garbage collector. arcpy could maintain connections to what are in reality unused variables in such a way as to fool the gc module.
As Nathan suggested you really need to post the code for a proper answer. We are all just guessing at this point...
But one think that may be causing you a problem is a memory leak with a cursor? In ArcGIS 10.1 you can use the search cursor with a WITH statement which guarantees its release. Have a look at the code sample here to see how to use it.
