On May 23, 2008, at 9:13 AM, jinzhu d wrote:


<excerpt>1
<fontfamily><param>Hiragino Kaku Gothic Pro</param>~{Hg:N@{SC~}</fontfamily>ublas<fontfamily><param>Hiragino Kaku Gothic Pro</param>~{Gs>X~}</fontfamily><fontfamily><param>STHeiti</param>~{Us5DDf#?~}</fontfamily>how
to get the inverse matrix ?

2
<fontfamily><param>Hiragino Kaku Gothic Pro</param>~{SP1H~}</fontfamily><fontfamily><param>STHeiti</param>~{=OMjU{5DVPND0oVzND<~#?~}</fontfamily>how
can i get a copy of helpfile in chinese ?


</excerpt>  Most of the time you don't need A^(-1) only a solution AX
= B or XA = B

which can be computed more efficiently than A^(-1)B or BA^(-1) using

a computed A^(-1).   for a general dense matrix would use an LU
decomposition

with partial pivoting.