BUG: ignore errors for invalid dates in to_datetime() with errors=coerce (#25512) #26561
Conversation
…as-dev#25512) parsing.try_parse_year_month_day() in _attempt_YYYYMMDD() throws not only ValueError but also OverFlowError for incorrect dates. So handling of this error was added.
Codecov Report
@@ Coverage Diff @@
## master #26561 +/- ##
==========================================
- Coverage 91.77% 91.76% -0.01%
==========================================
Files 174 174
Lines 50649 50649
==========================================
- Hits 46483 46479 -4
- Misses 4166 4170 +4
Continue to review full report at Codecov.
|
Codecov Report
@@ Coverage Diff @@
## master #26561 +/- ##
==========================================
- Coverage 91.77% 91.77% -0.01%
==========================================
Files 174 174
Lines 50649 50681 +32
==========================================
+ Hits 46483 46510 +27
- Misses 4166 4171 +5
Continue to review full report at Codecov.
|
| @@ -83,6 +83,35 @@ def test_to_datetime_format_YYYYMMDD(self, cache): | |||
| result = to_datetime(s, format='%Y%m%d', cache=cache) | |||
| assert_series_equal(result, expected) | |||
|
|
|||
| # GH 25512 | |||
There was a problem hiding this comment.
can you make a new test (and _overflow) in the name
| def test_to_datetime_format_YYYYMMDD_overflow(self, cache): | ||
| # GH 25512 | ||
| # NaN before strings with invalid date values, errors=coerce | ||
| s = Series(['19801222', np.nan, '20010012', '10019999']) |
There was a problem hiding this comment.
Could you parameterize over these Series inputs?
There was a problem hiding this comment.
Is it what you meant?
Thanks!
|
thanks @nathalier very nice. note that this might close another existing issue or 2, though searching for issues like this are hard |
git diff upstream/master -u -- "*.py" | flake8 --diffparsing.try_parse_year_month_day() in _attempt_YYYYMMDD() throws not only ValueError but also OverFlowError for incorrect dates. So handling of this error was added.