[SciPy-user] nonlinear fit with non uniform error?
Thu Jun 21 23:11:29 CDT 2007
Robert Kern wrote:
> Christian K wrote:
>> Is that the way scipy.odr handles the weights?
> The weighted sum of the residuals is the value that is minimized. I like
> formulating the problem that way rather than going through the details of the
> linear case. I think it's more direct this way.
I see. So then the same limitation applies: the error has to be gaussian, right?
More information about the SciPy-user