contribution-analysis

What problem does it solve?

Contribution analysis quantifies how much each factor contributes to explaining the variance of a response variable using the R² decomposition method. This skill enables precise attribution of predictive power to individual variables or factor groups, improving model interpretation and decision-making.

Core Features & Use Cases

• Global PCA-based factor scores: Combine variables, standardize, and derive factor scores for downstream contribution analysis.
• R²-based attribution: Compute each factor's share of explained variance via R² decomposition.
• Use Case: Data scientists can compare the influence of different feature groups on a target outcome to guide feature engineering and model selection.

Quick Start

Run a complete contribution analysis by fitting a global PCA across all variables, obtaining factor scores, and calculating each factor's contribution to the total R².