Rivu Basu Simulation Lab

Regression as a living experiment.

Build datasets, bend the fit, test predictions, and learn by doing. This upgraded lab turns a single regression demo into a visual diagnostic studio for residual behavior, outlier flags, leverage, and influential points.

Interactive canvas Residual diagnostics Cook's distance Influence ranking Linear + quadratic fit

Simulation Stage

y = 0.000 + 0.000x Prediction disabled Comparison hidden
data points fitted model confidence band comparison line residuals mean point prediction residual outlier influential point multiple flags
Points
0
R²
0.000
RMSE
0.000
Correlation
0.000

Regression Diagnostic Plots

Awaiting fit
Residuals vs Fitted
Normal Q-Q
Scale-Location
Cook's Distance

Regression Diagnostic Stats

Model fit
Adjusted R²
0.000
SSE
0.000
MAE
0.000
AIC
0.000
BIC
0.000
F statistic
0.000

Diagnostics Console

Awaiting fit
Intercept
0.000
Slope
0.000
Quadratic Term
0.000
Outliers Flagged
0
Mean Point
(0.00, 0.00)
Compare R²
0.000
No fit yet.

Walkthrough + Interpretation

Watching
What the model sees

The simulation is waiting for a pattern.

Calculation steps
Turn on walkthrough mode or fit a model to reveal the step-by-step calculations.
Teaching prompts

Influential Point Ranking

Top points
Point x y Std residual Leverage Cook's D Flag

Outlier Calculation

Rules
Decision rules
Fit a model to calculate standardized residuals, leverage, and Cook's distance.
Residual bands
Rivu Basu Simulation Lab is designed for hands-on statistical intuition: experiment first, diagnose second, explain with evidence third.