BUG: DateOffset with non-integer days is not working anymore

[SergejKr]

Pandas version checks

• I have checked that this issue has not already been reported.

• I have confirmed this bug exists on the latest version of pandas.

• I have confirmed this bug exists on the main branch of pandas.

Reproducible Example

import pandas as pd 
from datetime import datetime

datetime(2020, 1, 1) + pd.DateOffset(days=1.5)
datetime(2020, 1, 1) + pd.DateOffset(hours=1.5)

Issue Description

After updating pandas from 1.5.3 to 2.1.2 the behaviour of DateOffset has changed. When executing datetime(2020, 1, 1) + pd.DateOffset(days=1.5) I obtain datetime(2020, 1, 2) rather than datetime(2020, 1, 2, 12) . In the latter case datetime(2020, 1, 1) + pd.DateOffset(hours=1.5) I obtain the expected datetime(2020, 1, 1, 1, 30).

Expected Behavior

The expected behaviour should either be an Exception, when executing pd.DateOffset(days=1.5) with a non-integer input or the datetime(2020, 1, 2, 12).

Installed Versions

INSTALLED VERSIONS

commit : a60ad39
python : 3.9.13.final.0
python-bits : 64
OS : Windows
OS-release : 10
Version : 10.0.19045
machine : AMD64
processor : Intel64 Family 6 Model 140 Stepping 1, GenuineIntel
byteorder : little
LC_ALL : None
LANG : None
LOCALE : de_DE.cp1252
pandas : 2.1.2
numpy : 1.26.1
pytz : 2023.3.post1
dateutil : 2.8.2
setuptools : 68.2.2
pip : 23.3.1
Cython : 3.0.5
pytest : 7.4.3
hypothesis : None
sphinx : 7.2.6
blosc : None
feather : None
xlsxwriter : 3.1.9
lxml.etree : None
html5lib : None
pymysql : None
psycopg2 : None
jinja2 : 3.1.2
IPython : 8.17.2
pandas_datareader : None
bs4 : 4.12.2
bottleneck : None
dataframe-api-compat: None
fastparquet : None
fsspec : None
gcsfs : None
matplotlib : 3.8.1
numba : 0.58.1
numexpr : 2.8.7
odfpy : None
openpyxl : 3.1.2
pandas_gbq : None
pyarrow : 14.0.1
pyreadstat : None
pyxlsb : None
s3fs : None
scipy : 1.11.1
sqlalchemy : None
tables : 3.9.1
tabulate : 0.9.0
xarray : None
xlrd : 2.0.1
zstandard : None
tzdata : 2023.3
qtpy : 2.4.1
pyqt5 : None

[jbrockmendel]

i'd expect that to raise in the constructor

[jrmylow]

bisecting finds this first cropped up in this PR: 53681.

Summary: The line that causes this behaviour is this: offsets.pyx#L1356. This is due to how the relativedelta from dateutil handles multiplication incorrectly. I'm raising a bug report with them.

Tracing down, there is some interesting behaviour starting with the following setup:

import pandas as pd
import datetime as dt
import dateutil

start = dt.datetime(2020, 1, 1)
offset_1 =dt.timedelta(days=1.5)
offset_2 = pd.DateOffset(days=1.5)
offset_3 = dateutil.relativedelta.relativedelta(days=1.5)

Testing each offset reveals that multiplying the relativedelta is the cause here:

>>> start + offset_1
datetime.datetime(2020, 1, 2, 12, 0)
>>> start + offset_2
Timestamp('2020-01-02 00:00:00')
>>> start + offset_3
datetime.datetime(2020, 1, 2, 12, 0)
>>> start + offset_3 * 1
datetime.datetime(2020, 1, 2, 0, 0)

[jrmylow]

Raised here dateutil/issues/1318