如果 Nan 在单独的行中，则将值替换为 Nan - Pandas

Question

I have a column in a df that contains numerous strings. I was to replace some of the strings with Nan but there's too many to list that can be used to replace these values. I have a separate column that does contain Nan values, which may be used to achieve this.

I want to replace specific strings in Value . Just imagine if this column contains 1000 different strings and I want to replace 500 with Nan . it would be inefficient to create a list with these unwanted strings and use it to replace with Nan

There is a separate column (X) that displays Nan values that can be used to replace rows in Value . So where X is Nan , replace the row in Value with Nan .

Is there an easier way to do this?

df = pd.DataFrame({        
    'Value' : ['B','A','X','Y','C','D','E','F','G','H','I'],
    'X' : ['A','A','A','A',np.nan,'A','A','A',np.nan,'A','A'],  
    })

df = df.loc[df['X'].eq(np.nan), df['Value']] = np.nan
print(df)

Intended Output:

   Value    X
0      B    A
1      A    A
2      X    A
3      Y    A
4      Nan  NaN
5      D    A
6      E    A
7      F    A
8      NaN  NaN
9      H    A
10     I    A

Answer 1

You want DataFrame.mask with Series.isna

df=df.mask(df['X'].isna())
print(df)

---

   Value    X
0      B    A
1      A    A
2      X    A
3      Y    A
4    NaN  NaN
5      D    A
6      E    A
7      F    A
8    NaN  NaN
9      H    A
10     I    A

---

Also you can use DataFrame.where with Series.notna

df=df.where(df['X'].notna())

Answer 2

We can do dropna + reindex

df=df.dropna().reindex(df.index)
   Value    X
0      B    A
1      A    A
2      X    A
3      Y    A
4    NaN  NaN
5      D    A
6      E    A
7      F    A
8    NaN  NaN
9      H    A
10     I    A