Closed ocefpaf closed 1 year ago
numpy >=1.23
is causing the segfault, investigating...
I narrowed it down to the spline call in:
y_no_nans
is an array like,
array([], shape=(13, 0), dtype=float64)
I'm not sure if that is the correct way to test this. Even though it was passing before it looks like spline should not work here.
Confirmed about numpy
, im ok with the <=
for now until we can investigate further or the segfault is fixed in numpy
.
Let's see if we can observe the same segfault from #84 in this PR. Something is quite odd there :-/