BUG: `to_datetime` inconsistant between Series & DataFrame with timestamps

[ldouteau]

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

df = pd.DataFrame({"datecol": [1, 2]})

# Passes with Series
pd.to_datetime(df["datecol"])
# Fails with DataFrame
pd.to_datetime(df)

Issue Description

Converting timestamps to datetime works with Series, but fails with DataFrame

Error with DataFrame:

Traceback (most recent call last):
  File "C:\Users\...\main.py", line 8, in <module>
    pd.to_datetime(df)
  File "C:\Users\...\.venv\Lib\site-packages\pandas\core\tools\datetimes.py", line 1075, in to_datetime
    result = _assemble_from_unit_mappings(arg, errors, utc)
             ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "C:\...\.venv\Lib\site-packages\pandas\core\tools\datetimes.py", line 1191, in _assemble_from_unit_mappings
    raise ValueError(
ValueError: to assemble mappings requires at least that [year, month, day] be specified: [day,month,year] is missing

Expected Behavior

to_datetime should have consistent behaviour between Series & DataFrame.

Installed Versions

INSTALLED VERSIONS

commit : 4665c10
python : 3.12.10
python-bits : 64
OS : Windows
OS-release : 11
Version : 10.0.22631
machine : AMD64
processor : Intel64 Family 6 Model 140 Stepping 1, GenuineIntel
byteorder : little
LC_ALL : None
LANG : None
LOCALE : English_United Kingdom.1252

pandas : 2.3.2
numpy : 2.3.3
pytz : 2025.2
dateutil : 2.9.0.post0
pip : None
Cython : None
sphinx : None
IPython : None
adbc-driver-postgresql: None
adbc-driver-sqlite : None
bs4 : None
blosc : None
bottleneck : None
dataframe-api-compat : None
fastparquet : None
fsspec : None
html5lib : None
hypothesis : None
gcsfs : None
jinja2 : None
lxml.etree : None
matplotlib : None
numba : None
numexpr : None
odfpy : None
openpyxl : None
pandas_gbq : None
psycopg2 : None
pymysql : None
pyarrow : None
pyreadstat : None
pytest : None
python-calamine : None
pyxlsb : None
s3fs : None
scipy : None
sqlalchemy : None
tables : None
tabulate : None
xarray : None
xlrd : None
xlsxwriter : None
zstandard : None
tzdata : 2025.2
qtpy : None
pyqt5 : None

[Alvaro-Kothe]

This is taken from the docs

If a DataFrame is provided, the method expects minimally the following columns: "year", "month", "day". The column “year” must be specified in 4-digit format.

If you want to use pd.to_datetime you should use it like this:

import pandas as pd

df = pd.DataFrame({"day": [1, 2], "month": [1, 1], "year": [1970, 1970]})
print(pd.to_datetime(df))
# prints
# 0   1970-01-01
# 1   1970-01-02
# dtype: datetime64[s]

I don't think that this is a bug, since it is the expected behaviour.

[ldouteau]

This is taken from the docs

Oh right , i completely missed it.

I don't think that this is a bug, since it is the expected behaviour.

I agree that the docs are very clear about this, hence not a bug (my bad for the report). But it feels weird that timestamps are supported when working on series & not on DataFrame.

[Alvaro-Kothe]

But it feels weird that timestamps are supported when working on series & not on DataFrame

The key difference is that a DataFrame can have multiple columns. Hence, you should select the timestamp column.