Re: pseudo inverse
- From: spellucci@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx (Peter Spellucci)
- Date: Mon, 16 Jan 2006 18:29:43 +0000 (UTC)
In article <pan.2006.01.16.14.06.20.400720@xxxxxxxxxxxxxx>,
Andreas Schulze <andreas.schulze@xxxxxxxxxxxxxx> writes:
>Dear all,
>
>do you know a c library to calculate a (pseudo)-inverse of a matrix?
>Can I do this with lapack?
>
>Best regards
>Andreas
you get the (Moore-Penrose-)pseudoinverse easily from the singular value
decomposition
A=U*S*V' U and S unitaruy and S diagoanl and of the same shape as A
=>
A^# = V*S^#*U' and diag(S^#_{ii] = 1/S_{ii} if S_ii not=0 and 0 otherwise)
and for this one there are several choices in c available.
for example
in http:/www.netlib.org/c/meschach
http://www.netlib.org/clapack
in "numerical recipes in c"
hth
peter
.
- Follow-Ups:
- Re: pseudo inverse
- From: Dave Dodson
- Re: pseudo inverse
- References:
- pseudo inverse
- From: Andreas Schulze
- pseudo inverse
- Prev by Date: Napier's Bones for other algorithms?
- Next by Date: Least squares fit to absolute value function
- Previous by thread: pseudo inverse
- Next by thread: Re: pseudo inverse
- Index(es):
Relevant Pages
|