Topic: Transforming Predictors and Outcomes

Yeo-Johnson Transformations for Signed Variables

Zero and Negative Outcomes When Logs Are Inappropriate

Worked Translations Across Common Transformation Models

Using Log1p for Predictors With Zeros

Square-Root Predictors for Diminishing Returns

Square-Root Transformations of Outcomes

Shifting Negative Predictors Before Logging

Smearing Estimates for Arithmetic-Mean Predictions

Reciprocal Predictors for Saturating Relationships

Reciprocal Transformations of Outcomes