[SciPy-User] Problems evaluating interpolated function at new points

Fri May 15 10:26:23 EDT 2015

Update.

I just checked: I am running Python 2.7.9 and Scipy 0.14.0. As noted by
Oleksandr I may have been using the Syntax recommended for a newer
version.  Can anybody suggest a workaround or the correct syntax for
version 0.14.0?

Also, there was a typo when I run the notebook last and copied code in my
post (xnew, ynew, instead of xnew1, ynew1).  I run again my code with
xnew1, ynew1 and I still get the errors.

Thanks
Matteo

On Thu, May 14, 2015 6:24 pm, Oleksandr Huziy wrote:
> Hi All:
>
>
> This is not the answer to Matteo's question but probably a report on
> unclear documentation here:
>
> http://docs.scipy.org/doc/scipy-0.14.0/reference/generated/scipy.interpol
> ate.interp2d.html
>
> I am confused by the returns part, it says
>
>
> """
> Returns:
>
>
> *values_x* : ndarray, shape xi.shape[:-1] + values.shape[ndim:]
>
>
> Interpolated values at input coordinates.
>
>
> """
>
>
> But in the examples, the returned value f is used as a callable...
>
>
> """
>
>
> Now use the obtained interpolation function and plot the result:
>
>>>>
>
>>>> xnew = np.arange(-5.01, 5.01, 1e-2)>>> ynew = np.arange(-5.01,
>>>> 5.01, 1e-2)>>> znew = f(xnew, ynew)>>> plt.plot(x, z[0, :], 'ro-',
>>>> xnew, znew[0, :], 'b-')>>> plt.show()
>
> """
>
>
> Probably returns section corresponds to this version of scipy, but the
> examples are from the later version, which has probably confused Matteo.
>
>
> Cheers
>
>
>
>
> 2015-05-14 17:15 GMT-04:00 Matteo Niccoli <matteo at mycarta.ca>:
>
>
>> I am relatively new to Scipy, at least when it comes to interpolation.
>> I
>> am trying to replicate some of the functionality of Matlab's interp2.
>>
>> I have a distance function map (available at this Google drive
>> location)
>> https://drive.google.com/open?id=0B6acq_amk5e3X0Q5UG1ya1VhSlE&authuser=
>> 0
>>
>>
>> Values are normalized in the range 0-1. Size is 200 rows by 300
>> columns.
>>
>> I can load it up with this code snippet:
>> import numpy as np dstnc1=np.load('dstnc.npy')
>>
>>
>> Coordinates are defined by the next snippet:
>> xmin = 0. xmax = 9000. ymin = 0. ymax = 6000. r1,c1 = dstnc1.shape x =
>> np.linspace(xmin,xmax,c1) y = np.linspace(ymin, ymax,r1)
>>
>> I have three map points defined by vectors xnew1, ynew1 with this
>> snippet:
>> xnew1=[3700.540199,3845.940199,3983.240199]
>> ynew1=[1782.8611,1769.862,1694.862]
>>
>> I check their location with respect to the distance map with this:
>> import matplotlib.pyplot as plt fig = plt.figure(figsize=(20, 16)) ax =
>> fig.add_subplot(1, 1, 1) plt.imshow(dstnc1,
>> cmap=my_cmap_r,vmin=0,vmax=0.3, extent=[0, 9000, 0, 6000],
>> origin='upper') plt.scatter(xnew1, ynew1, s=50, linewidths=0.15)
>> plt.show()
>>
>> They plot in the correct location. Now I would like to extract the
>> distance value at those three points. I tried first interp2d.
>>
>> from scipy.interpolate import interp2d x1 = np.linspace(xmin,xmax,c1) y1
>> = np.linspace(ymin,ymax,r1)
>> f = interp2d(x1, y1, dstnc1, kind='cubic')
>>
>> but when I try to evaluate with: test=f(xnew,ynew)
>>
>> I get this error message:
>> ------------------------------------------------------------------------
>> ---
>> ValueError                                Traceback (most recent call
>> last) <ipython-input-299-d0f42e609b23> in <module>()
>> ----> 1 test=f(xnew,ynew)
>>
>>
>>
>> C:\...\AppData\Local\Continuum\Anaconda\lib\site-packages\scipy\interpo
>> late\interpolate.pyc in __call__(self, x, y, dx, dy) 270
>> (self.y_min, self.y_max)))
>> 271
>> --> 272         z = fitpack.bisplev(x, y, self.tck, dx, dy)
>> 273         z = atleast_2d(z)
>> 274         z = transpose(z)
>>
>>
>>
>> C:\...\AppData\Local\Continuum\Anaconda\lib\site-packages\scipy\interpo
>> late\fitpack.pyc in bisplev(x, y, tck, dx, dy) 1027     z,ier =
>> _fitpack._bispev(tx,ty,c,kx,ky,x,y,dx,dy)
>> 1028     if ier == 10:
>> -> 1029         raise ValueError("Invalid input data")
>> 1030     if ier:
>> 1031         raise TypeError("An error occurred")
>>
>>
>> ValueError: Invalid input data
>>
>>
>> If I try RectBivariateSpline:
>> from scipy.interpolate import RectBivariateSpline x2 =
>> np.linspace(xmin,xmax,r1) y2 = np.linspace(ymin,ymax,c1) f =
>> RectBivariateSpline(x2, y2, dstnc1)
>>
>>
>> I get this error:
>>
>>
>> -----------------------------------------------------------------------
>> ----
>> ValueError                                Traceback (most recent call
>> last) <ipython-input-302-d0f42e609b23> in <module>()
>> ----> 1 test=f(xnew,ynew)
>>
>>
>>
>> C:\...\AppData\Local\Continuum\Anaconda\lib\site-packages\scipy\interpo
>> late\fitpack2.pyc in __call__(self, x, y, mth, dx, dy, grid) 643
>> z,ier = dfitpack.bispev(tx,ty,c,kx,ky,x,y) 644                 if not
>> ier == 0: --> 645                     raise ValueError("Error code
>> returned by bispev: %s" % ier)
>> 646         else:
>> 647             # standard Numpy broadcasting
>>
>>
>> ValueError: Error code returned by bispev: 10
>>
>>
>>
>> Any suggestion as to whether I am using the wrong functions or the
>> right function with wrong syntax, and how I may fix it is appreciated.
>> Thank you
>>
>>
>>
>>
>> _______________________________________________
>> SciPy-User mailing list
>> SciPy-User at scipy.org
>> http://mail.scipy.org/mailman/listinfo/scipy-user
>>
>>
>
>
>
> --
> Sasha
> _______________________________________________
> SciPy-User mailing list
> SciPy-User at scipy.org
> http://mail.scipy.org/mailman/listinfo/scipy-user
>
>