Closed jmp75 closed 5 years ago
Changed to cast the raster.nodata
to a np.float32
. Works better with a raster where the band data is indeed returned as an np.float32 but remains to be seen whether works with other, preexisting cases e.g. and foremost on the Peel.
Another version of this issue is likely to show up on other data sets, but this can be closed.
Was getting very large magnitude AHD heights after blending COndamine DEM
Diagnosing in debug mode, it looks like a change of value from np.float32 to np.float64 on assigning (as an aside, unintuitive)
so, further comparison in whether v[i] is equal to raster.nodata fails.