Matrix Inverse/LAPACK

[Thanh Dang Nguyen <nguyet1@Eng.Auburn.EDU>]

Hi folks,

> Vi` he^. tho^'ng ddie^`u khie^?n to^i ddang thu+.c hie^.n la` o+? che^' ddo^. 
> Real-time ne^n Matlab kho^ng thich ho+.p. To^i muo^'n vie^'t la.i chuong trinh 
> tinh Inverse matrix ( ba^.c 3 va` ba^.c 6) tren C nhung quen mat thuat toan tinh. 

Since many people have answered to this post, I just want to add a useful
pointer that might be helpful for all of you who are involved in numerical 
linear algebra:

		http://www.netlib.org/lapack/

BTW, I guess you'd better reuse "standard", ie. well-written, well-documented
and well-tested library routines, rather than "reinventing the wheel" by
yourself. LAPACK has become a sort of standard package in numerical analysis,
and although the code you can download at the given site is not optimized
for specific hardware structures, their calling sequences have been 
standardized, so there would be no portability problems later on if you
want to use other commercial packages. All the source codes are free,
and available in Fortran, C/C++, and probably they have precompiled
libraries for different platforms, too.

After all these if you still would like to write the code by yourself,
then any introductory book on linear algebra should explain basic
methods for solving a system of linear equations, such as Gaussian
elimination, or QR algorithm, or other iterative methods.

Hope this helps!

-Thanh