Applyafunctiontoeachcellofapandasdataframeusinginformationfromaparticularcolumn

I want to expand the list entries of a dataframe using the information in column i:

i   s_1         s_1        s_3
2   [1, 2, 3]   [3, 4, 5]  NaN
1   NaN         [0, 0, 0]  [2]

The i value just indicates how often the last value of each list should be copied:

i   s_1               s_1              s_3
2   [1, 2, 3, 3, 3]   [3, 4, 5, 5, 5]  NaN
1   NaN               [0, 0, 0, 0]     [2, 2]

I am currently using a nested apply loop:

test.apply(lambda x: x.apply(
     lambda y: np.pad(y, (0, x.i), 'constant', constant_values=y[-1]) if type(y)==list else 0), axis=1)

However, this is very slow and if i have a lot of rows (>10.000) the code breaks. This solution seems a bit messy and i'm wondering what the best approach would be to do something like that?