What does print((a + b).tolist()) print? pandas Missing Data, NaN and NA

Answer, B. [2, <NA>, <NA>]

A. [<NA>, <NA>, <NA>]

Assumes a single missing value anywhere in either array poisons every position of the result rather than only the ones it actually touches.

B. [2, <NA>, <NA>]

Correct. Elementwise addition of two nullable Int64 arrays only produces a real value where both sides are present, so the first position gives 2, but the second and third each have one missing side and come back as <NA>.

C. [2, 2, <NA>]

Assumes a real value paired with a missing one on the other array still produces a real number instead of propagating the missing marker.

D. [2, 2, 3]

Assumes the missing slot on either side of the addition is simply ignored, as though it contributed nothing rather than making the result missing.

Elementwise arithmetic between two nullable arrays follows pd.NA propagation position by position, a missing value on either side of a given position makes that one result missing, and it has no effect on the other positions. This is different from a whole series reduction like sum(), where skipna controls whether a single missing value anywhere poisons the entire aggregate.