Modify the last row of each group in pandas MultiIndex DataFrame

I have a sample multi index dataframe below (actual much larger). I want to set the last row of each level = 0 index in the column 'col' to 100. Dataframe like this

mux = pd.MultiIndex.from_arrays([
list('aaabbbcccddd'),
list('tuvwacdjpkqz')],
names=['level 0', 'level 1'])
df = pd.DataFrame({'col': np.arange(len(mux))}, mux)
df

So the 2,5,8 and 11 wold all be replaced with 100. I have tried these solution but I dont know how to set the value properly. I can get the boolean series and the dataframe of the last rows but cant seem to figure out how to set them to the new value. What I have tried below

df.groupby(level=0).tail(1).col = 100
df.groupby(level=0).tail(1).col
df.iloc[-1, df.columns.get_loc('col')] = 100

The last line of code was from a previous post but it only sets the last row of df, not at the index level

Solution

You were close, use loc with the index of your result and assign:

df.loc[df.groupby(level=0).tail(1).index, 'col'] = 100
df

                 col
level 0 level 1     
a       t          0
        u          1
        v        100
b       w          3
        a          4
        c        100
c       d          6
        j          7
        p        100
d       k          9
        q         10
        z        100

This is your groupby result:

df.groupby(level=0).tail(1)
 
                 col
level 0 level 1     
a       v          2
b       c          5
c       p          8
d       z         11

Get the indices of the rows you want to change with .index:

df.groupby(level=0).tail(1).index

MultiIndex([('a', 'v'),
            ('b', 'c'),
            ('c', 'p'),
            ('d', 'z')],
           names=['level 0', 'level 1'])

You have index labels, so you will need loc (not iloc) for the final assignment step.