Hi Johann, I believe the scipy optimization package gives a variety of optimization algorithms but you would have to implement your own codes to fix/thaw parameters. That being said, there is a lot there and I haven't dug into all of it into detail. I would suggest posting a message to the scipy mailing list after the holidays and you may get a better response.
Cheers, Jessica On Dec 26, 2007, at 3:52 PM, Johann Cohen-Tanugi wrote: > hi jessica, thanks.... So scipy.optmizer as it stands cannot do > that? I gues I should move the issue to the scipy list then. Yes > there is mpfit, there is also pyminuit in google.code that is > wrapper of the high energy physics standard package MINUIT, etc.... > but I would think that fitting data, with fixed/thawed parameters, > is a basic feature of a scientific package. As there seems to be a > snergy between matplotlib and scipy, I would expect the latter to > take care of optimization.... > thanks for the replies, > Johann > > Jessica Lu wrote: >> Hi Johann, >> >> I would recommend using the python mpfit module: >> >> http://cars9.uchicago.edu/software/python/mpfit.html >> >> Cheers, >> Jessica >> >> On Dec 22, 2007, at 8:57 PM, Johann Cohen-Tanugi wrote: >> >>> hi jessica, >>> This FittingData tutorial is very nice. Could you illustrate how to >>> fix/thaw parameters? >>> I did not find any such attribute and when I try some kludges >>> they fail >>> with a msg saying >>> <type 'exceptions.ValueError'>: shape mismatch: objects cannot be >>> broadcast to a single shape >>> >>> >>> thanks, >>> Johann >>> >>> -------------------------------------------------------------------- >>> ----- >>> This SF.net email is sponsored by: Microsoft >>> Defy all challenges. Microsoft(R) Visual Studio 2005. >>> http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ >>> _______________________________________________ >>> Matplotlib-users mailing list >>> Matplotlib-users@lists.sourceforge.net >>> https://lists.sourceforge.net/lists/listinfo/matplotlib-users >>> >> > ------------------------------------------------------------------------- This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2005. http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ _______________________________________________ Matplotlib-users mailing list Matplotlib-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/matplotlib-users
