Re: [eigen] cache-friendly matrix inverse

[ Thread Index | Date Index | More lists.tuxfamily.org/eigen Archives ]


-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA256

Robert Lupton the Good schrieb:
> I'm not sure that this is still on-topic, but if you have a model
> [...] least-squares estimate [...]
> 
> So you really need to invert a matrix.  You can invert it any
> way you like (e.g. eigen-value decomposition; Gaussian elimination; ...)
> but invert it you must.

Well, I knew that you don't need to calculate an explicit inverse there
(it would have surprised me that such a standard task would invalidate
the no-inversion-in-numerics rule).

You need a SVD in this case. Anyway, as soon as you've got a numerical
task to solve, it's allways a good idea to have a look at the numerical
recepies book, in this case Chapter 15.4:

    http://www.fizyka.umk.pl/nrbook/c15-4.pdf


This leaves the question unanswered if these few posts are off-topic?
I guess: no. It's just a discussion about a usecase of Eigen ;)

CU,
Chris

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.9 (GNU/Linux)

iEYEAREIAAYFAkoLJ1cACgkQoWM1JLkHou1b+QCdGqYpzy+lzgEaUg08qme8sh4J
aD4An3mJB0O53bspef1WSYVCLRmNvUHQ
=oDBM
-----END PGP SIGNATURE-----



Mail converted by MHonArc 2.6.19+ http://listengine.tuxfamily.org/