The LAPACK forum has moved to https://github.com/Reference-LAPACK/lapack/discussions.

DGELSx family of functions: overdetermined matrix eqns

Open discussion regarding features, bugs, issues, vendors, etc.

DGELSx family of functions: overdetermined matrix eqns

Postby RRLapackuser » Fri Sep 21, 2007 7:34 pm

I'm failry new to LAPACK and have a quick question on the DGELSx family of functions for solving an overdetermined set of equations (Ax=b). It's a linear least squares (LLS) problem.

For my case, matrix A has more rows than columns.
b and x are column vectors

I'd like to solve them only if the rank of A = #columns in A
I also want to know the error (residual) in the solution (for full rank)

DGELS doesn't give me info about rank although error norm info is available (overwritten on B)

DGELSY gives me the rank. But is there a way to find out the error in the solution?

DGELSS seems like it might work for me. It gives rank and error info (overwritten on B) but it actually gives me a solution that is very off. Besides it computes SVD which I don't need.

Any suggestions for which function I should use?
RRLapackuser
 
Posts: 1
Joined: Fri Sep 21, 2007 7:01 pm

Return to User Discussion

Who is online

Users browsing this forum: No registered users and 4 guests