Filling in missing values in Python

SimpleImputer replaces gaps with a statistic learned from the training data.

import numpy as np
from sklearn.impute import SimpleImputer

train = np.array([[1.0, 10.0], [3.0, np.nan], [5.0, 30.0]])
test = np.array([[np.nan, 20.0]])

imputer = SimpleImputer(strategy="median")
filled = imputer.fit_transform(train)

print(imputer.statistics_)
print(filled)
print(imputer.transform(test))

How it works

1. strategy picks the mean, median or most frequent value.
2. It learns the fill value in fit, so the test set gets the same one.
3. statistics_ shows what it will substitute per column.

Keywords and builtins used here

• as
• print

The run, in numbers

Lines

12

Characters to type

309

Tokens

91

Three-star pace

105 tpm

At the three-star pace of 105 tokens a minute, this run takes about 52 seconds.

Type this snippet

Step 2 of 4 in Preprocessing, step 5 of 25 in Machine learning with scikit-learn.

← Previous Next →