If condition, store previous df column value, without loop

Question

is it possible to retrieve previous column value without loop in pandas?

import pandas as pd
import numpy as np

df=pd.DataFrame({'a': [True, False, False, True, True, False], 'b': [0, 0, 0, 3, 4, 4]})

df
Out[427]: 
       a  b
0   True  0
1  False  0
2  False  0
3   True  3
4   True  4
5  False  4

If df['a'] = True, store index, else previous value

# Attempt 1
df['c'] = np.where(df['a'], df.index, df['c'].shift(1))


# Attempt 2
df['c'] = df.index
df['c'] = np.where(df['a'], df['c'], df['c'].shift(1))

Answer 1

You can use a ffill on the index converted to_series after masking the non-True values:

df['b'] = df.index.to_series().where(df['a']).ffill(downcast='infer')

output:

       a  b
0   True  0
1  False  0
2  False  0
3   True  3
4   True  4
5  False  4