BUG: cut with non-nano #56101
BUG: cut with non-nano #56101
Conversation
mroeschke
added
cut
pandas/tests/reshape/test_cut.py
Outdated
| # TODO: constructing DatetimeIndex with dtype="M8[s]" without truncating | ||
| # the first entry here raises in array_to_datetime. Should truncate | ||
| # instead of raising? | ||
| # TODO: should we truncate to 7 seconds or 8? |
There was a problem hiding this comment.
If these are inclusive, left bin edge and in nanos the result was 7.2, I think I would expect 7 instead of 8
There was a problem hiding this comment.
looking at this closer, i think it doesn't really matter bc bins[0] is calculated roughly on min(data) - .001 * (max(data) - min(data)). that .001 * [...] adjustment looks intended to make sure that the bottom bin is a strict lower bound. Exactly how strict doesn't seem to matter. In this test case, max(data) - min(data) is a 2-day Timedelta, and when that Timedelta has second unit, multiplying it by .001 truncated to Timedelta(minutes=2, seconds=52), truncating the .8 seconds.
There was a problem hiding this comment.
Ah I forgot about the 0.001 offset applied to the edges (though it seem code-smelly). The current truncation result should be OK then
|
Thanks @jbrockmendel |
doc/source/whatsnew/vX.X.X.rstfile if fixing a bug or adding a new feature.sits on top of #56099