Open jay-hennen opened 6 years ago
Is it worth looking at the scipy code and submitting a patch?
( at the very least, a doc patch )
You'd think delta could be positive or negative, with no change in the code anywhere else...
I also see this:
"The data must be defined on a regular grid; the grid spacing however may be uneven."
Which is weird, as that isn't a "regular" grid -- it is a "rectangular" grid, though...
Can thank scipy.RegularGridInterpolator for this one. It only accepts dimensions for the data that are exclusively ascending rather than in sorted order (ascending or descending). This is not mentioned in it's documentation.
One possible workaround: detect and
numpy.flip
any offending dimensions, and flip that dimension's data as well. There are lazy-loading concerns with flipping the data dimension though.