The only errors that are currently supported are errors from a jackknife. This is slow, and in the case of correlated fits, may not be necessary on first pass.
I need to implement classical errors which come from the curvature of the chisq. This is explained here:
The only errors that are currently supported are errors from a jackknife. This is slow, and in the case of correlated fits, may not be necessary on first pass.
I need to implement classical errors which come from the curvature of the chisq. This is explained here:
http://www.mathworks.com/matlabcentral/newsreader/view_thread/157530