calculator for linear regression

Linear Regression Calculator

Enter your data as two equal-length lists of numbers. This tool computes the best-fit line y = mx + b using ordinary least squares.

Tip: You need at least 2 paired points, and all x values cannot be identical.

What this calculator does

This calculator estimates a straight-line relationship between one independent variable (x) and one dependent variable (y). In statistics, this is called simple linear regression. The output gives you the slope, intercept, correlation, and goodness-of-fit metrics so you can quickly evaluate how well a line explains your data.

How to use it

  • Enter all x values in the first field.
  • Enter all y values in the second field in the same order.
  • Optionally enter an x value for prediction.
  • Click Calculate Regression.

Example: If x is advertising spend and y is sales, the regression line helps estimate expected sales at future spending levels.

Understanding the output

1) Regression equation: y = mx + b

The calculator returns the best-fit line. The slope (m) is how much y changes for each 1-unit increase in x. The intercept (b) is the predicted value of y when x = 0.

2) Correlation (r)

Correlation ranges from -1 to +1 and measures direction and strength of linear association. Values near +1 indicate strong positive linear relationships, near -1 indicate strong negative linear relationships, and near 0 indicate weak linear association.

3) Coefficient of determination (R²)

R² indicates the fraction of variability in y explained by x through a linear model. For example, R² = 0.84 means 84% of variation in y is explained by the fitted line.

4) Residual sum of squares (SSE) and standard error

SSE captures total squared error between observed and predicted y values. Lower SSE generally means a tighter fit. Standard error estimates typical prediction error magnitude in y-units.

The core formulas

The calculator uses standard ordinary least squares formulas:

  • m = Sxy / Sxx
  • b = ȳ - m x̄
  • r = Sxy / √(Sxx · Syy)
  • R² = 1 - SSE/SST (when SST > 0)

where Sxx = Σ(x - x̄)², Syy = Σ(y - ȳ)², and Sxy = Σ(x - x̄)(y - ȳ).

Best practices and assumptions

  • Use this for roughly linear relationships.
  • Watch out for outliers; one extreme point can strongly affect slope.
  • Avoid extrapolating far beyond your observed x range.
  • Remember: correlation does not imply causation.

When this tool is useful

  • Trend estimation for business or operations data
  • Academic homework checks for statistics and data science courses
  • Quick forecasting from small datasets
  • Sanity checks before building more advanced predictive models

If your data has multiple predictors, non-linear patterns, or strong seasonality, consider moving to multiple regression, polynomial regression, or time-series models.

🔗 Related Calculators