set_feature_names

Set names for all features in the model.

Method call format

set_feature_names(feature_names)

Parameters

feature_names

Description

A one-dimensional array of feature names for each feature. The order and number of specified names must match the ones used in the dataset.

Possible types

  • ndarray
  • list

Default value

Required parameter.

Example

The following example outputs the original feature names, changes them and then outputs the updated names.

from catboost import CatBoost

train_data = [[1, 4, 5, 6],
              [4, 5, 6, 7],
              [30, 40, 50, 60]]

eval_data = [[2, 4, 6, 8],
             [1, 4, 50, 60]]

train_labels = [10, 20, 30]

model = CatBoost()

model.fit(train_data,
          train_labels,
          verbose=False)

print("Original feature names:")
print(model.feature_names_)
print("Changed feature names:")
model.set_feature_names(["feature_1", "feature_2", "feature_3", "feature_4"])
print(model.feature_names_)

Output:

Original feature names:
['0', '1', '2', '3']
Changed feature names:
['feature_1', 'feature_2', 'feature_3', 'feature_4']