Discussion
Up to the Feature Requests Forum
Please feel free to communicate anything you feel could help this project move on. You can receive an email each time a new message is posted by setting up your personal preferences.
scalability
Posted by
obadmin
at
February 21. 2007
Hi aaron,
well I don't think that OpenBayes has ever been used with such a large network and dataset. However, I trust that python will manage the memory space correctly. You should have lot's of memory on your PC in order to avoid using your hard disks swap which would make your PC run forever.
I invite you to give it a try and share with us your observations. You could probably first give it a try with a very large dummy network (containing the same node 1000 times for example) if you don't want to spend time creating your network.
This seems like a very interesting test to me. I'm waiting for your comments
Kosta
well I don't think that OpenBayes has ever been used with such a large network and dataset. However, I trust that python will manage the memory space correctly. You should have lot's of memory on your PC in order to avoid using your hard disks swap which would make your PC run forever.
I invite you to give it a try and share with us your observations. You could probably first give it a try with a very large dummy network (containing the same node 1000 times for example) if you don't want to spend time creating your network.
This seems like a very interesting test to me. I'm waiting for your comments
Kosta
Powered by Ploneboard
Version 0.2 available soon
I am a newbie to OpenBayes....
my question is to do with scalability in learning & inferencing. I have a very complex dataset consisting of
>50,000 nodes etc and I was curious if anyone has done experiments using OB on that scale. How did it scale? what kind of performance benchmarks did you achieve etc.
Thanks for sharing.