Closed jaycedowell closed 10 months ago
@dentalfloss1 @league Any opinions on this PR which would allow lower precision values to be stored in a "standard header"? I'm currently leaning towards keeping the current behavior in master
and not merging this one.
I agree.
Patch and project coverage have no change.
Comparison is base (
becbf48
) 65.39% compared to head (778c666
) 65.39%. Report is 2 commits behind head on master.
:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Have feedback on the report? Share it here.
Motion carries, closing.
It looks like this has sorta kinda already been merged in but with
numpy.int64
replacingnumpy.int
andnumpy.float64
replacingnumpy.float
. I think those are the correct sizes for native Python int's and float's so maybe this PR isn't needed? I guess the advantage of this approach is thatinteger
allows bothint32
andint64
, andfloating
worth withfloat32
andfloat64
.