debug:scikit-learn

What problem does it solve?

This Skill helps diagnose and fix common scikit-learn debugging issues in ML workflows, including NotFittedError, shape mismatches between training and test data, NaN or infinity values, and pipeline/configuration problems that hinder model performance.

Core Features & Use Cases

• NotFittedError identification and guidance: Detects when a model is used before fitting and provides practical steps to train before evaluation.
• Shape alignment and preprocessing validation: Verifies train/test feature compatibility and ensures consistent preprocessing with Pipelines and ColumnTransformer.
• NaN/Inf data quality checks and imputation advice: Detects missing or infinite values and suggests appropriate imputation strategies.
• Pipeline integrity and cross-validation debugging: Validates each pipeline step, flags misconfigurations, and helps diagnose leakage or over/underfitting across CV schemes.

Quick Start

Provide your X_train, y_train, X_test, y_test, and the scikit-learn pipeline to receive diagnostics and recommended fixes.