Bug: Don't intercept TypeError in DataFrame.agg

[rhshadrach]

pandas/pandas/core/apply.py

Lines 735 to 745 in 3370c81

	try:
	result = super().agg()
	except TypeError as err:
	exc = TypeError(
	"DataFrame constructor called with "
	f"incompatible data and dtype: {err}"
	)
	raise exc from err
	finally:
	self.obj = obj
	self.axis = axis

We intercept any TypeError and reword the message to be about the DataFrame constructor. This likely made sense before, but with the code being refactored and #43741, TypeErrors can occur for many different reasons. We should just remove the except block above.

[sarvaSanjay]

So just to confirm all we want to do is delete lines 737-742 ? Or do we have to add more specific ways to intercept type errors?

[rhshadrach]

So just to confirm all we want to do is delete lines 737-742 ?

That's correct; since this can raise for e.g. a user-defined function, I don't believe we can be any more helpful than just allowing the normal error to raise.

[sarvaSanjay]

take

[alexanderwertman3]

I would like to work on this for a school project

[yuvanist]

@sarvaSanjay If you are not updating the test for this, I'll take it. Please let me know.

[yuvanist]

take

[yuvanist]

@rhshadrach #49969 PR closes this. Please review.