CODEWITHSUNDEEP
pythonpython-3.xpandasdataframe
khemedi
khemedi

Reputation: 806

Why cannot change the value for a specific column in a slice of pandas data frame using .loc?

df1 = pd.DataFrame(np.random.random_sample((10)), columns=list('A'))
df1['MATCHED'] = 0

df1

          A  MATCHED
0  0.424651        0
1  0.855567        0
2  0.983395        0
3  0.921866        0
4  0.001827        0
5  0.341491        0
6  0.055578        0
7  0.970564        0
8  0.078751        0
9  0.348055        0

Then I filter df1:

df1_slice = df1[df1['A'] <=0.4]
df1_slice
          A  MATCHED
4  0.001827        0
5  0.341491        0
6  0.055578        0
8  0.078751        0
9  0.348055        0

Now, I want to change the column MATCHED values for those rows in df1_slice :

df1.loc[df1_slice.index]['MATCHED']=1

I'd expect the MATCHED column changes in df1 from 0 to 1, but it doesn't.

df1.loc[df1_slice.index]
          A  MATCHED
4  0.001827        0
5  0.341491        0
6  0.055578        0
8  0.078751        0
9  0.348055        0

Why they don't change and how to change this script so that they change to 1 in df1.

Upvotes: 0

Views: 30

Answers (1)

dkapitan
dkapitan

Reputation: 921

I find that using .loc throughout is the best way to (re-)assign to slices. In your example you use .loc only to slice on the index, not on the columns. This should do it (haven't tested):

df1.loc[df1_slice.index, 'MATCHED'] = 1

Upvotes: 1

Related Questions