[SciPy-User] scipy.optimize leastsq with numpy longdouble?

Joan Smith joans@MIT....
Wed Nov 17 13:10:06 CST 2010


Thank you, with those suggestions it worked!

Joan
On Nov 17, 2010, at 12:50 PM, Charles R Harris wrote:

> 
> 
> On Wed, Nov 17, 2010 at 10:25 AM, Joan Smith <joans@mit.edu> wrote:
> Hi,
> 
> What do you mean where is the data coming from? It's from an experiment..
> function I'm fitting is:
> maxwell_boltzmann = lambda v,x: v[0]*(x-C0)**(-4)*np.exp(-(L/v[1])**2*(x-C0)**(-2)) + v[2]
> 
> and the data is in an array of longdoubles.
> Does that answer your question?
> Joan
> 
> 
> That's a bit vague. Is the data of actual longdouble precision? 19 digits would be unusual accuracy for experimental data. Does the data have outliers? How much data do you have? The pdf doesn't look like the usual Maxwell-Boltzmann, where does it come from? What are L, C0, v, and x in the function?
> 
> Is there any reason you can't convert the experimental data to ordinary doubles?
> 
> On Nov 17, 2010, at 12:20 PM, Charles R Harris wrote:
> 
>> 
>> 
>> On Wed, Nov 17, 2010 at 10:13 AM, Joan Smith <joans@mit.edu> wrote:
>> Hi,
>> 
>> I'm fitting a maxwell-botlzmann distribution, using SciPy leastsq. Because of the data I'm fitting, I need high precision (on both the high and low ends), so I'm using numpy.longdouble for my types. Is there a way to use this type with leastsq? As it stands, I'm getting this error:
>> 
>>    v = leastsq(self.chi_squared, self.v_0,  args=self.args, full_output=1)
>>  File "/Library/Frameworks/Python.framework/Versions/2.6/lib/python2.6/site-packages/scipy/optimize/minpack.py", line 281, in leastsq
>>    maxfev, epsfcn, factor, diag)
>> TypeError: array cannot be safely cast to required type
>> 
>> I've been using leastsq for a while, and haven't seen this error before, so I suspect it has to do with using an unusual type.
>> 
>> 
>> Yep. But where is your data coming from? And how are you trying to do the fit? I suspect that a change of method it the proper way to go here. Josef may have something in the stats model package.
>> 
> 
> 
> Chuck
> 
> _______________________________________________
> SciPy-User mailing list
> SciPy-User@scipy.org
> http://mail.scipy.org/mailman/listinfo/scipy-user

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mail.scipy.org/pipermail/scipy-user/attachments/20101117/b30ae960/attachment.html 


More information about the SciPy-User mailing list