scipy.interpolate.interpnd complaining about 'Delaunay' object has no attribute 'simplices'

Question

I'm dusting off some code I wrote a few months ago, and for some reason it doesn't work anymore... In a nutshell, I'm using scipy.interpolate.LinearNDInterpolator objects to interpolate models and compare to data. Now, when I attempt to call the interpolator object with the coordinates at which I would like the interpolation, I get the following error:

In [9]: a([[3500, 3.5, 1.5]])
AttributeError                            Traceback (most recent call last)
<ipython-input-9-91f2103e7a0c> in <module>()
----> 1 a([[3500, 3.5, 1.5]])

/usr/lib64/python2.7/site-packages/scipy/interpolate/interpnd.so in     scipy.interpolate.interpnd.NDInterpolatorBase.__call__ (scipy/interpolate/interpnd.c:3133)()

/usr/lib64/python2.7/site-packages/scipy/interpolate/interpnd.so in     scipy.interpolate.interpnd.LinearNDInterpolator._evaluate_double (scipy/interpolate/interpnd.c:3954)()

/usr/lib64/python2.7/site-packages/scipy/interpolate/interpnd.so in scipy.interpolate.interpnd.LinearNDInterpolator._do_evaluate (scipy/interpolate/interpnd.c:4684)()

AttributeError: 'Delaunay' object has no attribute 'simplices'

I have never seen this error before, and the code has worked previously. Did something just change in scipy that I'm not aware of?

Thanks for looking!

Answer 1

I guess you use an older version of the library:

The Delaunay library has two different accessors for simplices: "Delaunay.simplices" and "Delaunay.vertices" shown here (newest docs): http://docs.scipy.org/doc/scipy/reference/generated/scipy.spatial.Delaunay.html

Of the two Delaunay.vertices is marked "deprecated".

On Ubuntu 13.04 the simplices call does not exist however, because it still uses scipy 0.11.0: http://docs.scipy.org/doc/scipy-0.11.0/reference/generated/scipy.spatial.Delaunay.html#scipy.spatial.Delaunay

Try with this minimal example or just rewrite your simplices call to vertices:

from __future__ import print_function

import numpy as np
from scipy.spatial import Delaunay
import sys

my_molecule = np.random.rand(400,3) #points for query
points = np.random.rand(1000, 3)   #points used for Triangulation

diag = Delaunay(points)
simplices = diag.find_simplex(my_molecule)

for point,simplex in zip(my_molecule,simplices):
    if simplex == -1:
        print ("Point not included in diag.")
        continue
    print ("Doing vertices call: ")
    spoints = diag.vertices[simplex]
    print ("Doing simplices call: ")
    spoints = diag.simplices[simplex]