%timeit df.apply(lambda row: row['sales'] * (1 - row['discount']), axis=1)