Type a unit name or symbol; use ↑ ↓ and Enter or Tab to accept a suggestion. Compound units work too: kN*m → ft*lbf.
One letter is a variable (F, IG, μ); a word searches by name (friction, inertia). Greek names and sqrt turn into symbols as you type; Tab undoes it. ↑ ↓ pick, Enter inserts.
Search a material (aluminum, 6061, stainless, water), a property (yield, density, specific heat) or a symbol (E, cp, ρ). Click a value to put it in your line in SI units, or + line to add it as its own variable. ↑ ↓ pick, Enter inserts the first value.
Pick a topic to see how it works. Try it opens the example in a new tab. New here?
The regression line passes through (x̄, ȳ) with slope r·s_y/s_x. Tabacus finds the slope, the intercept and a prediction, and has R² and residuals too.
Least-squares line from summary statistics: ŷ = a + b·x Given: correlation r = 0.8, s_x = 2, s_y = 5, x̄ = 10, ȳ = 60 Slope: b = r · s_y / s_x = 0.8 · 5 / 2 = 2 Intercept: a = ȳ − b · x̄ = 60 − 2 · 10 = 40 Line: ŷ = 40 + 2x Prediction at x = 12: ŷ = 40 + 2 · 12 = 64 R² = r² = 0.64, so the line explains 64% of the variation in y.
The least-squares intercept is a = ȳ − b·x̄, so plugging in x = x̄ always gives ŷ = ȳ.
The actual y minus the predicted ŷ. Least squares picks the line that makes the sum of squared residuals as small as possible.
The calculator above is already loaded with this problem, edit it or type your own.