[SciPy-User] SciPy-User Digest, Vol 106, Issue 6

Nadav Horesh nadavh@visionsense....
Fri Jun 8 01:50:20 CDT 2012


The default interpolation method in linear. Does you case falls in between (cubic --- overshoot, linear --- too coarse)?
________________________________________

2012/5/31 Nadav Horesh <nadavh@visionsense.com>:
> You can try also scipy.interpolate.griddata (not the pylab's griddata!)
>
> ? Nadav
>
> _______________________________________________
> SciPy-User mailing list
> SciPy-User@scipy.org
> http://mail.scipy.org/mailman/listinfo/scipy-user
>

I tried this function and it works very similar to the interp2
function of matlab. The only problem is that my data are very peaky
and the cubic version of griddata gives me big over/under-shoots. The
matlab function does not gives me such behavior.

Is there another high 2D order interpolator that could be used in this case ?

Octave has a nice "pchip" option for interp2 that use piecewise cubic
Hermite interpolating polynomial, but unfortunatly griddata does not
have such option.

Thanks

Christine


More information about the SciPy-User mailing list