BUG: `groupby.agg` with `as_index=False` and duplicate columns in `by` fails with `pd.NamedAgg`. about pandas

[rhshadrach]

Thanks for the report. You checked the box I have confirmed this bug exists on the main branch of pandas. I am seeing the following output on main:

print(df.groupby(by=["col1", "col1", "col2"], as_index=False).agg(new_col=pd.NamedAgg('col1', min), new_col1=pd.NamedAgg('col1', max), new_col2=('col2', 'count')))
#    col1  col2  new_col  new_col1  new_col2
# 0     0     5        0         0         1
# 1     0     7        0         0         1
# 2     1     5        1         1         1
# 3     1    36        1         1         1
# 4     2     4        2         2         2

Can you confirm if you are seeing an error or the above output on main?

[sfc-gh-rdurrani]

I can't seem to install pandas from source - I keep getting the following error:

ModuleNotFoundError: No module named 'pandas._libs.pandas_parser'

but I did notice that the bug is happening on the latest version of pandas - 2.2.2.

[VISWESWARAN1998]

@sfc-gh-rdurrani Your code runs fine without any errors in main branch for me.

I can't seem to install pandas from source - I keep getting the following error:

ModuleNotFoundError: No module named 'pandas._libs.pandas_parser'

but I did notice that the bug is happening on the latest version of pandas - 2.2.2.

Try creating an isolated environment using mamba (recommended) or pip, install the requirements in requirements-dev.txt and try installing pandas from source. More information here: https://pandas.pydata.org/docs/dev/development/contributing_environment.html