[SciPy-user] scipy.signal.chebwin

[Kumar Appaiah]

On Fri, Feb 15, 2008 at 02:33:05PM -0600, Ryan May wrote:
> > I am unable to figure out where the problem could be, though I guess
> > it would have to do with the Chebyshev polynomial evaluation. I could
> > really do with a little help in debugging the chebwin fix. :-)
> > 
> Got it!
> 
> It seems scipy.special.chebyt doesn't quite do what we want it to do.
> If I replace the call to chebyt with this function:
> 
> def myT(order, x):
>     retval = N.cosh(order*N.arccosh(N.abs(x)))
>     retval[N.abs(x)<=1] = N.cos(order*N.arccos(x[N.abs(x)<=1]))
>     return retval
> 
> I get the correct window values.  I had checked plots of chebyt and the
> analytic definition, and they looked the same, but the actual values
> differed.  For most values they're close, but not identical.  Anybody
> know the difference between the Chebyshev polynomial defined as above
> and the one in scipy.special.chebyt?

I owe you a $BEVERAGE (non-alcoholic). Could you please help me in
getting the fix in by updating the patch?

Thanks.

Kumar
-- 
Kumar Appaiah,
458, Jamuna Hostel,
Indian Institute of Technology Madras,
Chennai - 600 036
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 189 bytes
Desc: Digital signature
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20080216/8e5c57c7/attachment.sig>