fit_transform() function learns from the data, does the encoding, and then, transforms the data.

Please note that the fit_transform() method returns <class ‘scipy.sparse._csr.csr_matrix’>. We can transform this matrix into an array and print it.

So, the output of the above program will be:

Value counts before filling missing values: 
Southampton    644
Cherbourg      168
Queenstown      77
Name: embark_town, dtype: int64
Value counts after filling missing values: 
Southampton    646
Cherbourg      168
Queenstown      77
Name: embark_town, dtype: int64
<class 'scipy.sparse._csr.csr_matrix'>
[[0. 0. 1.]
 [1. 0. 0.]
 [0. 0. 1.]
 ...
 [0. 0. 1.]
 [1. 0. 0.]
 [0. 1. 0.]]

Now, let’s say we want to add these values as three different columns in the dataset. We can do so in the following way:

import seaborn
from sklearn.preprocessing import OneHotEncoder

df = seaborn.load_dataset("titanic")

df["embark_town"].fillna(value="Southampton", inplace=True)

one_hot_encoder = OneHotEncoder()
transformed_df = one_hot_encoder.fit_transform(df[["embark_town"]])
print(type(transformed_df))
print(transformed_df.toarray())
print(one_hot_encoder.categories_[0])
df[one_hot_encoder.categories_[0]] = transformed_df.toarray()
print(df.head())

Now, print(one_hot_encoder.categories_[0]) will print the following: …