I have a general question on nonlinear least squares fitting using the Levenberg Marquardt algorithm. I would like to be able to handle data sets that are correlated—have a non-diagonal covariance matrix. Thus the weighting matrix would also be nondiagonal, but the vi in labview only allows the weights to be a vector of 1/sigma^2 (i.e., no correlation). Has anyone modified the necessary vis to handle a weighting matrix rather than a vector? If not, I could try it myself, but how do I copy the vis to make the changes? The Save As is always grayed out.
This VI is "buried" inside an internal LabVIEW library, which may explain why you can't "Save As". However, you can Select All (^A), Copy, open a new VI, Paste, and save your new VI. Edit away ...
