Ridge & Lasso

Ridge & Lasso is a fast, fully offline app that fits regularized linear regression - Ridge (L2) and Lasso (L1) - and shows exactly how the coefficients shrink as the penalty grows.

The signature view is the coefficient-shrinkage path: every predictor's coefficient sliding toward zero as the penalty lambda increases, with Lasso snapping weak predictors to exactly zero while Ridge only shrinks them close.

WHAT'S INSIDE

Paths
The signature shrinkage-path chart - every predictor's coefficient plotted against the penalty, with the current lambda marked and a plain-language verdict.

Data
A predictors x outcome table (add or remove columns and rows) with the fitted standardized and raw coefficients, intercept, R-squared, and the active-predictor count.

Concept
A plain-English walkthrough of why we shrink coefficients, Ridge vs Lasso, and the penalized loss functions - with formulas rendered in offline KaTeX.

Ridge vs Lasso
A side-by-side look at the two methods - L2 shrinks every coefficient a little, L1 can zero some out entirely, doubling as automatic feature selection.

Simulate
Knobs for signal strength, noise, and how many predictors are truly relevant versus pure noise - great for watching Lasso pick out the real signal.

Examples
Ready-made multi-predictor datasets - house prices, car mileage, student scores, and crop yield - that load with one tap.

WHY RIDGE & LASSO

- 100% offline - your data never leaves your device
- No account, no sign-up, no clutter
- Clean, distinctive interface built for students, researchers, and anyone learning regularization
- Ridge solved in closed form, Lasso by coordinate descent - the real algorithms, not a toy
- Input is never lost - your whole design, including method and lambda, is saved automatically

Whether you're a student learning why we penalize big coefficients, a researcher deciding between Ridge and Lasso for a real dataset, or just curious how feature selection works, Ridge & Lasso shows you the whole mechanism in one glance.