[SciPy-user] using optimize.leastsq with numarray in python 2.3

aurelien.gourrier at free.fr aurelien.gourrier at free.fr
Wed Apr 6 09:50:38 CDT 2005


Dear all,

First of all, as I'm new to this list, hi ! I discovered SciPy about 2 weeks ago
and joined the list and I'm quite impressed with the level of some of the
discussions... I myself am quite a beginner in the field of scientific
programming and am learning a lot by following the discussions eventhough I
don't get it all...

So far, I have been trying some features, in particular, I'm very interested in
the least-square fitting routine of the optimize function. I tried it for very
simple functions and a reduced number of params and it seems to do a very good
job. However, with some data, I just couldn't get the right fit...

In this case, I read a simple data file into a list, transformed it into an
array (numarray) with type = float32. After quite some fighting wondering what
was wrong with my data, I found out that this format (i.e. explicitely
specifying float32) induces a pb in the result. I am aware of some possible
conflicts when using numarray, but still haven't got a clue as why, does anyone
know ?

Thanks in advance,

Cheers,

Aurélien

PS : Also, is it possible to use scipy under python 2.4 ?




More information about the SciPy-user mailing list