Check values greater than int with condition Pandas

Say, I have this dataframe:

t = pd.DataFrame(data={"Pos": [1, 2, 3], "A": [10, 2, 90], "B":[90, 98, 10]})

I would like to know which of the (values in A and B >= 20) but only if (Pos = 1 or 3).

The result should look like this:

"A"     "B"
False   True
True    True
True    False

I can find the values >= 20 using t[["A", "B"]].ge(20) but I cannot seem to find a way to add the condition based on Pos.

Would a kind soul be able to help?

Answer

You can use a mask:

t[['A', 'B']].ge(20).where(t['Pos'].isin([1, 3]), True)

Alternative with numpy.logical_or and boolean inversion ~:

np.logical_or(t[["A", "B"]].ge(20), ~t['Pos'].isin([1, 3]).values[:,None])

output:

       A      B
0  False   True
1   True   True
2   True  False