BUG: Fix resample with np.timedelta64 loffset has no effect #18708
Conversation
|
Hello @Licht-T! Thanks for updating the PR. Cheers ! There are no PEP8 issues in this Pull Request. 🍻 Comment last updated on December 10, 2017 at 02:51 Hours UTC |
Licht-T
force-pushed
the
fix-numpy-timedelta64-loffset-resample
branch
from
3f2ccbc to
77cba2b
Compare
| @@ -1051,6 +1051,8 @@ def __init__(self, freq='Min', closed=None, label=None, how='mean', | |||
|
|
|||
| if isinstance(loffset, compat.string_types): | |||
| loffset = to_offset(loffset) | |||
| elif isinstance(loffset, np.timedelta64): | |||
There was a problem hiding this comment.
much better to do the conversion up front
Timedelta already handles these conversions (timedelta and np.timedelta)
offsets should continue to use to_offset
as those are converted to absolute days for Timedelta
| loffset=timedelta(minutes=1)).mean() | ||
| # GH7687 | ||
| result1 = s.resample('5min', closed='right', label='right', | ||
| loffset=np.timedelta64(1, 'm')).mean() |
Codecov Report
@@ Coverage Diff @@
## master #18708 +/- ##
==========================================
- Coverage 91.6% 91.56% -0.05%
==========================================
Files 153 153
Lines 51273 51275 +2
==========================================
- Hits 46970 46949 -21
- Misses 4303 4326 +23
Continue to review full report at Codecov.
|
jreback
added
Reshaping
|
can you also test for some invalid offsets |
|
closing as stale. if you want to continue working, pls ping. |
git diff upstream/master -u -- "*.py" | flake8 --diff