Re: [SciPy-user] on global fitting for sets of data