groupby.quantile() fails on Timedelta columns

[dokteurwho]

Code Sample, groupby quantile example

# groupby quantile example
import pandas as pd

print(f"Pandas {pd.__version__}")

df = pd.DataFrame({
    'idx': [1, 1, 2],
    'start_date': ['2019-10-02', '2019-10-04', '2019-10-05'],
    'arrival_date': ['2019-11-02', '2019-10-14', '2019-10-15']})

for c in ['start_date', 'arrival_date']:
    df[c] = pd.to_datetime(df[c])

df['delta'] = df['arrival_date'] - df['start_date']

print(df.groupby('idx')[['delta']].quantile(.9))

Problem description

The code example used to work in 0.21.1.
In 0.25 I obtain a "TypeError: No matching signature found"

Expected Output

Pandas 0.21.1

           delta

idx
1 28 days 21:36:00
2 10 days 00:00:00

Output of pd.show_versions()

INSTALLED VERSIONS

commit : None
python : 3.6.8.final.0
python-bits : 64
OS : Linux
OS-release : 3.10.0-693.el7.x86_64
machine : x86_64
processor : x86_64
byteorder : little
LC_ALL : None
LANG : en_US.UTF-8
LOCALE : en_US.UTF-8

pandas : 0.25.3
numpy : 1.17.3
pytz : 2019.3
dateutil : 2.8.1
pip : 18.1
setuptools : 40.8.0.post20191016
Cython : 0.29.14
pytest : None
hypothesis : None
sphinx : None
blosc : None
feather : None
xlsxwriter : None
lxml.etree : None
html5lib : None
pymysql : None
psycopg2 : None
jinja2 : None
IPython : None
pandas_datareader: None
bs4 : None
bottleneck : None
fastparquet : None
gcsfs : None
lxml.etree : None
matplotlib : None
numexpr : None
odfpy : None
openpyxl : None
pandas_gbq : None
pyarrow : None
pytables : None
s3fs : None
scipy : None
sqlalchemy : None
tables : None
xarray : None
xlrd : None
xlwt : None
xlsxwriter : None

[WillAyd]

The method was overhauled in 0.25 to push it down from the Python space to C. If you want to poke around the implementation is in pandas.core.groupby.groupby

[jreback]

a related version (different error on master)

In [8]: x = pd.DataFrame({'time': pd.date_range('20200101', periods=10, tz='UTC'), 'value': pd.to_timedelta(np.arange(10), unit='s')}).set_index('time')             

In [9]: x                                                                                                                                                            
Out[9]: 
                                    value
time                                     
2020-01-01 00:00:00+00:00 0 days 00:00:00
2020-01-02 00:00:00+00:00 0 days 00:00:01
2020-01-03 00:00:00+00:00 0 days 00:00:02
2020-01-04 00:00:00+00:00 0 days 00:00:03
2020-01-05 00:00:00+00:00 0 days 00:00:04
2020-01-06 00:00:00+00:00 0 days 00:00:05
2020-01-07 00:00:00+00:00 0 days 00:00:06
2020-01-08 00:00:00+00:00 0 days 00:00:07
2020-01-09 00:00:00+00:00 0 days 00:00:08
2020-01-10 00:00:00+00:00 0 days 00:00:09

In [10]: x.resample('2D').quantile(0.99)                                                                                                                             
TypeError: Cannot cast TimedeltaArray to dtype float64