Replacing -inf values to np.nan in a feature pandas.series [duplicate]

问题

This question already has answers here:

Pandas replacing elements not working (3 answers)

Closed last year.

I want to replace the -inf values in a pandas.series feature (column of my dataframe) to np.nan, but I could not make it.

I have tried:

    df[feature] = df[feature].replace(-np.infty, np.nan)
    df[feature] = df[feature].replace(-np.inf, np.nan)
    df[feature] = df[feature].replace('-inf', np.nan)
    df[feature] = df[feature].replace(float('-inf'), np.nan)

But it does not work. Any ideas how to replace these values?

Edit:

df[feature] = df[feature].replace(-np.inf, np.nan)

works

BUT:

df =  df.replace(-np.inf, np.nan)

does not work.

回答1:

The problem may be that you are not assigning back to the original series.

Note that pd.Series.replace is not an in-place operation by default. The below code is a minimal example.

df = pd.DataFrame({'feature': [1, 2, -np.inf, 3, 4]})

df['feature'] = df['feature'].replace(-np.inf, np.nan)

print(df)

#    feature
# 0      1.0
# 1      2.0
# 2      NaN
# 3      3.0
# 4      4.0

回答2:

it should work:

df.replace([np.inf, -np.inf], np.nan,inplace=True)

来源：https://stackoverflow.com/questions/49814077/replacing-inf-values-to-np-nan-in-a-feature-pandas-series