can any one send me the codining for finding out inverse of a 6x6 matrix
inverse of 6x6 matrix
Collapse
X
-
Tags: None
-
Feel free to post some progress you have made on the problem.
Please include where your difficulty lies.
If you wish to find the formula to find the inverse (as opposed to a c program that works this out) I suggest you try a Math forum (Dr Math (search for him online) is pretty good in my experience) -
actually i need mathematical formulation which is used for finding inverse of matrix
minimum of 6x6 orderComment
-
Originally posted by r035198xThen a maths forum is best for you
For exact inverses, you'll want to look for "Gaussian elimination with partial pivoting". (There's also Gaussian elimination with full pivoting, but I've rarely found that necessary.)
However, I have to ask--do you actually need the full inverse of the matrix, or are you just trying to solve Ax = b? I ask because it is computationally more efficient to just calculate x without finding the full inverse of A, even if you're solving the system repeatedly. Some people ask for the full inverse of the matrix, when in reality they only want to solve a particular linear system.
Do you plan on mostly doing small matrices, on the order of 6 x 6 to maybe 20 x 20? Or do you eventually plan larger systems? And again, do you plan upon solving for the actual inverse, or do you only need to solve linear systems? In that case, iterative techniques are your friend. (Such as biconjugate gradient, successive over-relaxation, Gauss-Seidel, Jacobi, etc.) -- PaulComment
-
i want to do full matrix inverse for minimum 6x6 it is ok for me
simillarly i want to find eigen values and eigenvectors of a 6x6 matrix
if u help me in this regard it will be too much helpful for me........Ajeet
Originally posted by macklin01For any matrix of nontrivial size, you're not going to find an explicit formula for the inverse. (Okay, you can with cofactor expansions, etc., but not terribly practical.)
For exact inverses, you'll want to look for "Gaussian elimination with partial pivoting". (There's also Gaussian elimination with full pivoting, but I've rarely found that necessary.)
However, I have to ask--do you actually need the full inverse of the matrix, or are you just trying to solve Ax = b? I ask because it is computationally more efficient to just calculate x without finding the full inverse of A, even if you're solving the system repeatedly. Some people ask for the full inverse of the matrix, when in reality they only want to solve a particular linear system.
Do you plan on mostly doing small matrices, on the order of 6 x 6 to maybe 20 x 20? Or do you eventually plan larger systems? And again, do you plan upon solving for the actual inverse, or do you only need to solve linear systems? In that case, iterative techniques are your friend. (Such as biconjugate gradient, successive over-relaxation, Gauss-Seidel, Jacobi, etc.) -- PaulComment
-
The GNU scientific library embeds a set of routines dealing with matrices.
Real symetric matrices inversion and eigen vectors computation are part of the offered features. The library is open source, so you can either use it as is, or take a peek at the source code.
Comment
Comment