only if you know its going to be col_B, we can also do something like this, for 'n' columns you can loop

if (df.columns.contains("col_B")) df else df.drop("col_B")