CODEWITHSUNDEEP
pythonpandasfillna
NSK
NSK

Reputation: 105

Conditional row-wise fillna when previous' row condition is also met

Suppose I have the following df

import pandas as pd
import numpy as np

test = pd.DataFrame(data = {
                    'a': [1,np.nan,np.nan,4,np.nan,5,6,7,8,np.nan,np.nan,6],
                     'b': [10,np.nan,np.nan,1,np.nan,1,1,np.nan,1,1,np.nan,1]
                     })

I would like to use pd.fillna(method='ffill') but only when two separate condition are both met:

  1. Both elements of a row are NaN
  2. The element of the previous row of column 'b' is 10

Note: the first row can never be NaN

I am looking for a smart way - maybe a lambda expression or a vectorized form, avoiding for loop or .iterrows()

Required result:

result= pd.DataFrame(data = {
                    'a': [1,1,1,4,np.nan,5,6,7,8,np.nan,np.nan,6],
                     'b': [10,10,10,1,np.nan,1,1,np.nan,1,1,np.nan,1]
                     })

Upvotes: 1

Views: 332

Answers (1)

jezrael
jezrael

Reputation: 862551

You can test if forward filled value in b is 10 and also if all columns are filled with missing values and pass to DataFrame.mask with ffill:

mask = test['b'].ffill().eq(10) & test.isna().all(axis=1)

test = test.mask(mask, test.ffill())
print (test)
      a     b
0   1.0  10.0
1   1.0  10.0
2   1.0  10.0
3   4.0   1.0
4   NaN   NaN
5   5.0   1.0
6   6.0   1.0
7   7.0   NaN
8   8.0   1.0
9   NaN   1.0
10  NaN   NaN
11  6.0   1.0

Upvotes: 4

Related Questions