TST: fix DecimalArray._reduce kludges #29630
Conversation
pandas/core/groupby/ops.py
Outdated
| # np.bool_(False) when we have all-NA DecimalArray | ||
| if group.isna().all(): | ||
| res = group.dtype.na_value | ||
| result[label] = res |
There was a problem hiding this comment.
It's not clear to why this is the correct fix. You are changing the result of the function, without necessarily knowing what the function is?
There was a problem hiding this comment.
yah, i want to take another shot at fixing this behavior inside DecimalArray.__array_whatever__
There was a problem hiding this comment.
tentatively looks like a fix in DecimalArray._reduce gets the job done, so the change here has been reverted. We'll see if the CI agrees.
| try: | ||
| return cls(scalars) | ||
| except TypeError: | ||
| # handling int 0 is necessary for groupby tests |
There was a problem hiding this comment.
Can you explain why this is necessary? What is a code example running into this?
There was a problem hiding this comment.
this is made unnecessary by the fix i just pushed, so has been reverted
jreback
added
Dtype Conversions
|
@jorisvandenbossche gentle ping |
* TST: fix DecimalArray._reduce kludges
* TST: fix DecimalArray._reduce kludges
No description provided.