contribution-analysis

You calculate factor contributions using R² decomposition, which attributes the explained variance of a response variable to individual factors. The process applies global PCA to derive factor scores and computes each group's share of predictive power.

R² decomposition is a statistical method that breaks down the explained variance of a response variable to attribute predictive power to individual correlated factors. It quantifies how much each feature group contributes to a target outcome.

Standardize your dataset and fit a global PCA across all variables to extract factor scores. These scores are then used in R² decomposition calculations to attribute explained variance to specific factor groups.

Yes, this workflow automates data standardization, global PCA fitting, factor score extraction, and R² contribution calculations. It directly outputs a structured result comparing the influence of different feature groups on a target outcome.

Use R² decomposition when you need to quantify the exact contribution of correlated variables to a response variable's variance. Standard factor analysis identifies latent structures, while R² decomposition directly attributes predictive power for model selection.

PCA-based contribution analysis requires data standardization and works best with multiple correlated variables. It attributes explained variance globally, so it may not capture non-linear relationships or interactions outside the principal component structure.