CODEWITHSUNDEEP
pythonpandas
jos97
jos97

Reputation: 405

How to replace directly NaN values in a column?

I have a dataframe like this:

article_id          title
NaN                 title_1
NaN                 title_2
NaN                 title_3
'202102011404103'   title_4
'202102011404104'   title_5
NaN                 title_6

Firstly I would like to add a condition 'if there is NaN value do the following' (else do nothing). If there is NaN value in the column: I would like to add directly an incremental value to NaN values in the column with something like this:

x = 1
df.insert(0, 'article_id', range(x, x + len(df)))

But I don't know how to add directly the code above to the column article_id, only to NaN values. How can I do it ?

Expected output:

article_id              title
    1                   title_1
    2                   title_2
    3                   title_3
    '202102011404103'   title_4
    '202102011404104'   title_5
    4                   title_6

Upvotes: 0

Views: 43

Answers (1)

jezrael
jezrael

Reputation: 863411

You can create mask for compare missing values and pass range with first value with count of NaNs by sum:

m = df['article_id'].isna()

x = 1
df.loc[m, 'article_id'] =  range(x, x + m.sum())
print (df)
          article_id    title
0                  1  title_1
1                  2  title_2
2                  3  title_3
3  '202102011404103'  title_4
4  '202102011404104'  title_5
5                  4  title_6

Upvotes: 2

Related Questions