Skip to Data Skip to Chart Skip to Results

Regression by EyeStatLens

Generate a random dataset with a linear trend. Each click produces new data with a different slope, spread, and sample size.

Choose a dataset:

Edit or paste two-column data (CSV with headers):

Open a CSV, TSV, or JSON file:

X variable (predictor): Y variable (response):

Drag the green line up/down, or drag its endpoints to change the slope.

Show residuals Show squared residuals Show best-fit line

Help

Drag the green line to fit the data as closely as you can. Use the controls to reveal residuals and compare your line to the mathematically optimal fit.

Grab the line: Grab near an end to pivot; grab near the middle to shift parallel
Show Residuals: Show vertical lines from each point to your line
Show Squared Residuals: Show literal squares whose area equals the squared error at each point
Absolute / Squared: Switch between Sum of Absolute Errors (SAE) and Sum of Squared Errors (SSE)
Show Best-Fit Line: Reveal the mathematically optimal line for the current metric

Keyboard Shortcuts

?: Show this help
↑ / ↓: Shift line up/down (parallel move)
← / →: Adjust slope (tilt left/right)
Shift+arrow: Move in larger steps
R: Try Again (reset your line)
N: Generate new random data
Esc: Close this dialog

New to this tool? Walk through it step by step.

Open guided walkthrough ↗