Re: [Help-gsl] Inverse of a matrix

[Ralph Silva]

Hi,

attached I have my own functions for multivariate normal, Student t and
Wishart distributions.  You can take a look and see how the inverse of a
matrix is calculated.

All the functions were compared with the results of R-2.2.1 software (
http://www.r-projec.org ) and seem to be working perfectly.

Regards,
Ralph.






On 4/29/06, 廖宮毅 <address@hidden> wrote:
> Thanks for attention, I have a question about computing the inverse of a
> matrix, for example, I want to calculate the value of the P.D.F. with a
> set of sample(size N) Y from a multivariate normal distribution
> (p-variate), with mean MU and covariance matrix SIGMA,  the P.D.F is :
>    det(SIGMA)^(0.5*N)*exp(-0.5*(Y - MU)%*%inv(SIGMA)%*%(Y -MU))
> my problem is: how to use GSL calculating the inverse of SIGMA?
> I have found some solutions on the mailing list, but the solutions are
> not easy to understand (due to my lack of knowledge), is there any
> suggestion computing the inverse of a matrix (in a clear and direct
> way)? Any comment is appreciated!
>
>
>
>
>
> _______________________________________________
> Help-gsl mailing list
> address@hidden
> http://lists.gnu.org/mailman/listinfo/help-gsl

rmv.c
Description: Text Data

rmv.h
Description: Text Data

test.c
Description: Text Data

Makefile
Description: Binary data