burcin -- what I did at Dev1

I wanted to make Sage compute nullspaces of matrices over polynomial rings in a reasonable time.

I started off with implementing fraction free elimination over ZZ[x] and realised I really needed fast polynomial arithmetic. So

I also fixed some bugs

I benchmarked the performance of Maple, MMA and Magma on real world example matrices. Maple seems to perform best overall. Custom Nullspace implementations in MMA can beat Maple for some special cases. Magma is not very good at this problem. I'll post detailed results soon.

I had a lot of discussions with Arne and Clement on how to go about solving this problem, and learned that

It is hard!

I completed implementing the fraction free elimination,

I also tried

Ideas for new algorithms:

dev1/burcin (last edited 2008-11-14 13:42:04 by anonymous)