BUG: Groupby any/all raises TypeError for pd.NA

[phofl]

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

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

• (optional) I have confirmed this bug exists on the master branch of pandas.

---

Note: Please read this guide detailing how to provide the necessary information for us to reproduce your bug.

Code Sample, a copy-pastable example

import pandas as pd

df = pd.DataFrame({"a": [1, 1, 2, 2, 3, 3, 4], "b": [1, pd.NA, 2, 3, 4, 5, 6]})
grouped = df.groupby("a")
print(grouped.all())
print(grouped.any())

Problem description

Both raise a TypeError

Traceback:

Traceback (most recent call last):
  File "/home/developer/.config/JetBrains/PyCharm2020.2/scratches/scratch_4.py", line 10, in <module>
    print(grouped.all())
  File "/home/developer/PycharmProjects/pandas/pandas/core/groupby/groupby.py", line 1424, in all
    return self._bool_agg("all", skipna)
  File "/home/developer/PycharmProjects/pandas/pandas/core/groupby/groupby.py", line 1375, in _bool_agg
    return self._get_cythonized_result(
  File "/home/developer/PycharmProjects/pandas/pandas/core/groupby/groupby.py", line 2631, in _get_cythonized_result
    raise TypeError(error_msg)
TypeError: boolean value of NA is ambiguous

Maybe using mask like in print(df.all(axis=1))?

Expected Output

Should work like with np.nan and return

      b
a      
1  True
2  True
3  True
4  True
      b
a      
1  True
2  True
3  True
4  True

Output of pd.show_versions()

master

[phofl]

Similar:

df = pd.DataFrame({"a": [1, 1, 2, 2, 3, 3, 4], "b": [1, pd.NA, 2, 3, 4, 5, 6]})
grouped = df.groupby("a")
print(grouped.mean())

raises a DataError

Traceback:

Traceback (most recent call last):
  File "/home/developer/.config/JetBrains/PyCharm2020.2/scratches/scratch_4.py", line 10, in <module>
    print(grouped.mean())
  File "/home/developer/PycharmProjects/pandas/pandas/core/groupby/groupby.py", line 1490, in mean
    return self._cython_agg_general(
  File "/home/developer/PycharmProjects/pandas/pandas/core/groupby/generic.py", line 1021, in _cython_agg_general
    agg_mgr = self._cython_agg_blocks(
  File "/home/developer/PycharmProjects/pandas/pandas/core/groupby/generic.py", line 1123, in _cython_agg_blocks
    raise DataError("No numeric types to aggregate")
pandas.core.base.DataError: No numeric types to aggregate

while

print(grouped.sum()) works.

[arw2019]

I think this falls under the umbrella of #37494

[arw2019]

I think this falls under the umbrella of #37494

On a second look I think this is an independent issue

[GYHHAHA]

The groupby.all or groupby.any problem comes from bool(pd.NA) is ambiguous.

pandas/pandas/core/groupby/groupby.py

Line 1367 in d378852

vals = np.array([bool(x) for x in vals])