在 pandas 数据框中的两列组合的行之间填充

Question

I have a data frame like this,

col1    col2    col3
 1        A      T
 2        A      F
 3        N      N
 4        N      N
 5        B      T
 6        N      N
 7        B      F
 8        N      N
 9        A      T
 10       N      N
 11       N      N
 12       A      T
 13       N      N
 14       N      N
 15       A      T
 16       N      N
 17       A      F

Now I want to create a new data frame from above in such a way that, of there is consecutive N in col2 and col3 between every T and F in col3, fill with above non N value with T. Ignore those where T or F coming after T and F and.

So the desired data frame look like,

col1    col2    col3
 1        A      T
 2        A      F
 3        N      N
 4        N      N
 5        B      T
 6        B      T
 7        B      F
 8        N      N
 9        A      T
 10       N      N
 11       N      N
 12       A      T
 13       N      N
 14       N      N
 15       A      T
 16       A      T
 17       A      F

I could do this using a for loop and store the indices by comparing the next and previous value. But it will take the longer time to execute. I am looking for some pythonic way/ pandas shortcuts to do it efficiently.

Answer 1

This is my approach:

# mask T and F
TFs = df['col3'].mask(df['col3'].eq('N'))

after_T = TFs.ffill()
before_F = TFs.bfill()

# between
bt_TF = after_T.eq('T') & before_F.eq('F')

# mask and ffill:
df['col2'] = df['col2'].mask(bt_TF).ffill()
df['col3'] = df['col3'].mask(bt_TF).ffill()

Output:

    col1 col2 col3
0      1    A    T
1      2    A    F
2      3    N    N
3      4    N    N
4      5    B    T
5      6    B    T
6      7    B    F
7      8    N    N
8      9    A    T
9     10    N    N
10    11    N    N
11    12    A    T
12    13    N    N
13    14    N    N
14    15    A    T
15    16    A    T
16    17    A    F