Implementation:Interpretml Interpret Clean Dimensions

Field	Value
Sources	InterpretML
Domains	Data_Preprocessing, Validation
Last Updated	2026-02-07 12:00 GMT

Overview

Clean_Dimensions is a concrete tool for validating and normalizing raw input data dimensions provided by the InterpretML library.

Description

The clean_dimensions function accepts any array-like input (lists, tuples, DataFrames, Series, masked arrays, sparse matrices) and converts it to a validated numpy ndarray with correct dimensionality. It handles edge cases like single-element arrays, masked arrays, and pandas types. This function is a key component of the data preparation pipeline, ensuring that all downstream operations receive data in a consistent and validated format.

Usage

Import this function when you need to validate raw user-provided y, sample_weight, or other 1D/2D data before feeding it into EBM training.

Code Reference

Source Location

Repository: interpretml/interpret
File: python/interpret-core/interpret/utils/_clean_simple.py
Lines: 49--230

Signature

def clean_dimensions(data, param_name):

Import

from interpret.utils._clean_simple import clean_dimensions

I/O Contract

Inputs

Name	Type	Required	Description
`data`	Any (array-like)	Yes	Raw input data to validate
`param_name`	str	Yes	Name used in error messages for debugging

Outputs

Name	Type	Description
return	`np.ndarray`	Cleaned numpy array with validated dimensions

Usage Examples

Basic Usage

import numpy as np
from interpret.utils._clean_simple import clean_dimensions

# Clean a list into a numpy array
y_raw = [1, 0, 1, 1, 0]
y_clean = clean_dimensions(y_raw, "y")
# y_clean is now np.array([1, 0, 1, 1, 0])

# Clean a pandas Series
import pandas as pd
weights = pd.Series([1.0, 2.0, 1.5])
weights_clean = clean_dimensions(weights, "sample_weight")

Related Pages

Page Connections

Double-click a node to navigate. Hold to expand connections.

Principle

Implementation

Heuristic

Environment