encode dataframe with python using encoder

Question

I 'am trying to encode a dataframe with LabelEncoder() before creating my machine learning model

here the code :

from sklearn.preprocessing import LabelEncoder
# LabelEncoder
le = LabelEncoder()

# apply "le.fit_transform"
df_encoded = data1.apply(le.fit_transform)
print(df_encoded)
print(le.classes_)

But I got this error :

TypeError: ("'<' not supported between instances of 'str' and 'NoneType'", 'occurred at index SACC_MARKET_SEGMENT')

Anyone can help ùme to resolve this problem? tahnk you

Rafaó · Accepted Answer · 2019-01-30 13:59:23Z

1

There can be a problem with type of your data. I don't know what's your desired data type, but you can try converting data1 to string:

from sklearn.preprocessing import LabelEncoder
le = LabelEncoder()

df_encoded = le.fit_transform(data1.astype(str))
print(df_encoded)
print(le.classes_)

answered Jan 30, 2019 at 13:59

Rafaó

5994 silver badges17 bronze badges

Sign up to request clarification or add additional context in comments.

4 Comments

Nasri Over a year ago

thank you for your answer, yes there are NaN missing values in this dataframe. I try you code but I got this error :` ValueError: bad input shape (266572, 11)`

Rafaó Over a year ago

that's because LabelEncoder needs 1D array. If you want to encode 2D array (like a DataFrame), try using OrdinalEncoder - works the same, but it's rather used to encode 2D stuff. All you have to do is take my code and replace LabelEncoder with OrdinalEncoder.

Nasri Over a year ago

Ok thanks, but I got this error: cannot import name 'OrdinalEncoder'

Rafaó Over a year ago

Maybe you have an old version of sklearn and you don't have OrdinalEncoder available yet. Check your version: print(sklearn.__version__). If it's smaller 0.20.2, then try to upgrade your sklearn package. If you use pip as package manager, try running this command in CMD: pip install scikit-learn --upgrade

Collectives™ on Stack Overflow

encode dataframe with python using encoder

1 Answer 1

4 Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

4 Comments

Your Answer

Sign up or log in

Post as a guest

Related