Linear Algebra: Colt/Jama vs Apache Commons Math 3

Edit: I wrote this while working through the example in

This was before I noticed OG-Maths and OG-Lapack, so I’m not sure if my observations below are relevant. Probably not.

Sorry. Still learning…

Hello,

I’ve only just learned about OpenGamma and I’m excited by what I see. I think it can help me a lot and I hope I will be able to contribute back.

Since you started in 2009 and I’m starting in 2014, the world is a little different now. For linear algebra, I see that you are using Colt and Jama. If I go to the official Colt website, it says it was last updated in 2004 and the Download link is broken. I can find the download page via Google, but the version is the same as the one on the official website.

There is a similar story for Jama. The latest version on the official website is from 2012 and there were no changes since then.

I was so excited about OpenGamma, I spent some time this morning revising my analytics, which uses Apache Commons Math 3, to OG analytics instead, but after digging into Colt/Jama a bit, I’m not sure that is a good idea.

I’m curious, how huge of an effort would it be to translate your analytics to Apache Commons Math 3?

Is this something you might be interested in?

It is just crazy enough, I may give it a try :smile:

Edit: I continue to dig through the code and think its great!

Best wishes,
Eric

Just noting that we have an effort to produce a high performance maths library - OG-Maths which will ultimately replace some of these libraries. Right now, I don’t think we would want to replace what we have with Commons Math 3.

1 Like