Re: covariance branch update

[John Darrington]

On Mon, May 17, 2010 at 04:51:11PM -0400, Jason Stover wrote:
     I just pushed a rewrite of linreg.c and regression.q
     to the covariance branch. It uses the code in covariance.c.
     Please let me know if looks OK. It passed make check.

Great!  Just one small thing I noticed:  You have a new static function invert_r
which isn't used.  Should this be deleted or do you have plans for it in the 
future?
     
     If it does look OK, then I should merge this to master
     and delete covariance-matrix.[ch], coefficient.[ch],
     and design-matrix.[ch].


Getting rid of those files will certainly be a good thing.   There will be
some conflicts when you try to merge, because src/math/covariance.[ch] 
changed a little when I added the FACTOR command.  Let me know if you have 
problems resolving these conflicts, and I'll see what I can do to help.
     
     The next step will be to make covariance.[ch] able
     to handle interactions. That could be difficult.
     
I think the bulk of the work is done there. src/math/interactions.c handles
the creation of interactions.  Creating a covariance matrix from an interaction
should not be too hard, I think (but perhaps I'm being naive).   Perhaps when 
we've merged we can talk about that.

On the subject of merging branches, I'm of the opinion that they're beginning to
proliferate.  An audit of the repository is needed to see which ones can be 
deleted or
merged.

J'

-- 
PGP Public key ID: 1024D/2DE827B3 
fingerprint = 8797 A26D 0854 2EAB 0285  A290 8A67 719C 2DE8 27B3
See http://pgp.mit.edu or any PGP keyserver for public key.

signature.asc
Description: Digital signature