[SciPy-User] fast N-d interpolators Intergrid and Barypol

denis denis-bz-py@t-online...
Tue Mar 12 12:14:08 CDT 2013

  two small fast interpolators:

interpolate in an N-d box grid, uniform or non-uniform.  
This is just a wrapper for scipy.ndimage.map_coordinates and numpy.interp.

interpolate in a uniform N-d box grid, using d + 1 corners of a simplex  
(triangle, tetrahedron ...) around each query point.  
>From Munos and Moore, "Variable Resolution Discretization in Optimal Control",  
1999, 24p; see the pictures and example on pp. 4-5.  
It's implemented in a C++ header barypol.h, with a Cython wrapper
(both of which more knowledgeable people could certainly improve).

In 4d, 5d, 6d, Intergrid does around 3M, 2M, .8M interpolations / second.  
Barypol is ~ 5 times faster, but not as smooth.  
(Both will of course become cache-bound for large grids.)  
See interpol/test/*.log
and http://github.com/denis-bz/interpol .

Comments are welcome, testcases most welcome.

  -- denis

More information about the SciPy-User mailing list