Matrix Inverse calculation problem.

help me with it.

[Geoff Hayes]

Rajeev - please describe (or show) what you have tried so far. Indicate which part of the code you are having problems with or not understanding.

Answer 1

[Mark Stone]

If you have a matrix A, and want to calculate the inverse of it using MATLAB, then inv(A) is the inverse of A. If this is not sufficient to answer your question, then you need to explain what you are trying to do.

[Walter Roberson]

inv(A) should not often be used. pinv() or the backslash operator are usually better choices.

[Bjorn Gustavsson]

Whenever pinv is necessary to use it is strongly advisable to not use it. When trying to invert very illconditioned matrices or solve very ill-determined systems of equations surely it is strongly preferable to have better grasp of what is going on than pinv provides - that is a full, or partial, SVD decomposition is the better choice to start the analysis.

[Mark Stone]

pinv is just the ticket for linear least squares. solution = pinv(A) * b , computed using rock solid SVD, much more reliable than A\b .

I'm not quite sure why anyone, including me, is humoring the O.P. given the question as posed. I chose not to enter into a discussion that, often or usually, the inverse isn't needed, and you should use \ or whatever. I was leaving that for round 2 when/if the O.P. asked a reasonable question.

[Bjorn Gustavsson]

Mark, I agree on all(most) all counts - the only objection I have is that when solving ill-conditioned inverse problems one should always look carefully at the eigenvalues and eigenvectors (and consider the full range of suitable regularizations).