|
From: | John W. Eaton |
Subject: | Re: Octave/C++ matrix Inv() comparison |
Date: | Tue, 09 Jul 2013 10:40:20 -0400 |
User-agent: | Mozilla/5.0 (X11; Linux x86_64; rv:10.0.12) Gecko/20130116 Icedove/10.0.12 |
On 07/09/2013 09:31 AM, ionone wrote:
Thanks you I tried both *dgetrf/dgetri* and *dsytrf/dsytri* to inverse my matrix but i still have differences between Lapack and Octave. For example in Octave i have : *1.7338e+010* and with *dgetrf/dgetri* i have : *17351976742.876453* and with *dsytrf/dsytri* i have : *17352023489.128361* so you see they differ quite a lot... Do you have an idea what is the Inv() function of Octave in Lapack's formaty ? i tried to look into the code but i couldn't find anything relevant.
You can start by reading the inv.cc file. I'm a bit surprised that it is hard to find that file and trace what the inv function does from there. Anyway, for dense real-valued matrices, it ultimately calls the Matrix::inverse function defined in dMatrix.cc.
In any case, this is really the wrong list for this question. The maintainers list is intended to be a place for discussion about the future development of Octave, not for questions about how to use Octave. Could you please move this discussion to the help list?
Thanks, jwe
[Prev in Thread] | Current Thread | [Next in Thread] |