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
import numpy as np

In [1]: pd.Timedelta(np.timedelta64(1, "M"))
Out[1]: Timedelta('31 days 00:00:00')

In [2]: np.timedelta64(1, "M").astype('timedelta64[D]')
Out[2]: numpy.timedelta64(30,'D')

Issue Description

A numpy numpy.timedelta64(1,'M') is converted by pandas.Timedelta as an interval of 31 days.

Expected Behavior

It should be converted to an interval of 30 days.

In pandas 1.5.x it was converted to Timedelta('30 days 10:29:06')

Installed Versions

INSTALLED VERSIONS

commit : 478d340
python : 3.11.1.final.0
python-bits : 64
OS : Darwin
OS-release : 22.4.0
Version : Darwin Kernel Version 22.4.0: Mon Mar 6 20:59:28 PST 2023; root:xnu-8796.101.5~3/RELEASE_ARM64_T6000
machine : arm64
processor : arm
byteorder : little
LC_ALL : None
LANG : None
LOCALE : None.UTF-8

pandas : 2.0.0
numpy : 1.24.2
pytz : 2023.3
dateutil : 2.8.2
setuptools : 65.5.0
pip : 22.3.1
Cython : None
pytest : None
hypothesis : None
sphinx : None
blosc : None
feather : None
xlsxwriter : None
lxml.etree : None
html5lib : None
pymysql : None
psycopg2 : None
jinja2 : None
IPython : 8.12.0
pandas_datareader: None
bs4 : None
bottleneck : None
brotli : None
fastparquet : None
fsspec : None
gcsfs : None
matplotlib : None
numba : None
numexpr : None
odfpy : None
openpyxl : None
pandas_gbq : None
pyarrow : None
pyreadstat : None
pyxlsb : None
s3fs : None
scipy : None
snappy : None
sqlalchemy : None
tables : None
tabulate : None
xarray : None
xlrd : None
zstandard : None
tzdata : 2023.3
qtpy : None
pyqt5 : None

In [1]: should probably raise a ValueError. The equivalent pd.Timedelta(1, unit="M") now raises a ValueError

.astype('timedelta64[D]') has a default casting='unsafe', the equivalent np.timedelta64(np.timedelta64(1, "M"), "D") would raise

In [3]: np.timedelta64(np.timedelta64(1, "M"), "D")
---------------------------------------------------------------------------
TypeError                                 Traceback (most recent call last)
Cell In[3], line 1
----> 1 np.timedelta64(np.timedelta64(1, "M"), "D")

TypeError: Cannot cast NumPy timedelta64 scalar from metadata [M] to [D] according to the rule 'same_kind'

Agreed with @mroeschke.

Side-note: I've been wondering recently if we should just map low-reso datetime64/timedelta64s to Period/DateOffsets. It's not great typing-wise, but seems right abstraction-wise.

Going to jump on this

take