in reply to Data Regression Utility
Just a suggestion -- you might consider writing your OLS in matrix notation, and then using PDL to sling the matrices, rather than handling all the messy algebra scalar-wise. The matrix approach would also give you multivariate OLS for free, vs. just the single dimensional y=mx+b.
|
---|
Replies are listed 'Best First'. | |
---|---|
Re^2: Data Regression Utility
by ~~David~~ (Hermit) on Sep 09, 2004 at 13:04 UTC |
In Section
Cool Uses for Perl