Function scipy.interpolate.interpn() 太慢了

Question

I am using scipy.interpolate.interpn() to interpolate a value at a single point on a 2D grid. The catch is that for this single point and for the 2D grid, I have to perform many evaluations of the interpn function:

interpolatedValue = []
for i in range(300000):
  interpolatedValue.append(scipy.interpolate.interpn(points, valueList[i], point)[0])

The code works as expected, but is far too slow. For each of these calls, the parameters point and points are the same. Consequently the interpolation weights determined for each call must be identical. Since the weights are the same, it is inefficient to compute them over and over again.

Can it really be that there is no faster way to do this with scipy?

Answer 1

interpolatedValue = [ scipy.interpolate.interpn( points, value, point)[0] for value in valueList ]