CODEWITHSUNDEEP
pythonarrayspandasnumpy
NRKirby
NRKirby

Reputation: 1634

Why can't I assign array to pandas DataFrame column?

I am using KNeighborsClassifier to classify some values like this:

arr = classifier_3NN.predict(testData_df)
len(arr)
10960

I want to assign this array to a column in a DataFrame, I have checked & it is the same size:

len(df[df['events']=='NaN']['events'])
10960

When I do the following command the array values are not in the column as expected:

df[df['events']=='NaN']['events'] = arr

Can anyone see what I'm doing wrong?

Upvotes: 2

Views: 2223

Answers (1)

jezrael
jezrael

Reputation: 862661

I think you need isnull for checking NaN values with ix:

df.ix[df['events'].isnull(), 'events'] = arr

But if need replace NaN values by arr better is use fillna by Series created from arr:

arr = [5,7,9]

df = pd.DataFrame({'events':[np.nan, 1, np.nan]})
print (df)
   events
0     NaN
1     1.0
2     NaN

df['events'] = df['events'].fillna(pd.Series(arr, index=df.index))
print (df)
   events
0     5.0
1     1.0
2     9.0

Upvotes: 1

Related Questions