BUG: series resample with timedelta values looses dtype (GH13119) #14118
Conversation
| @@ -89,7 +89,8 @@ def trans(x): # noqa | |||
|
|
|||
| if issubclass(dtype.type, np.floating): | |||
| return result.astype(dtype) | |||
| elif dtype == np.bool_ or issubclass(dtype.type, np.integer): | |||
| elif (dtype == np.bool_ or issubclass(dtype.type, np.integer) and | |||
There was a problem hiding this comment.
change these to use is_bool_dtype and such (iow thr pandas intro spectors)
There was a problem hiding this comment.
ah yes, this probably solves the timedelta problem at once as well
There was a problem hiding this comment.
have to be careful about timedeltas because they are represented as integers, so new to be clear which dtype you have (and don't mix); this is more of an in general statement (not necessarily relevant to this particular question)
There was a problem hiding this comment.
The issue in this case was exactly something like that: a timedelta dtype passed the issubclass(dtype.type, np.integer) check, and for this reason the specific elif close for datetimes and timedeltas more below at line 128 was not used for timedeltas (but was correct for datetimes).
Current coverage is 85.27% (diff: 100%)@@ master #14118 diff @@
==========================================
Files 139 139
Lines 50511 50511
Methods 0 0
Messages 0 0
Branches 0 0
==========================================
+ Hits 43071 43072 +1
+ Misses 7440 7439 -1
Partials 0 0
|
git diff upstream/master | flake8 --diff