Curve Fitting & Regression Calculator

Regression Model

Model type

Polynomial degree (1–10)

Fit options

Force through origin Show residuals

Paste data (one pair per line)

Accepts: x, y or x<TAB>y or x y per line.
Lines starting with # are ignored.

Or load a sample dataset

Target language

Function name

Horner's method Use Horner's form (fewer multiplications — recommended for MCU)

Horner's form evaluates aₙ·xⁿ + … + a₀ as (…((aₙ·x + aₙ₋₁)·x + aₙ₋₂)…)·x + a₀, reducing n² multiplications to n for degree-n polynomial.

X min

X max

Number of entries (4–1024)

Run a fit to see the equation…

Copy this function into your firmware. No floating-point library needed for integer-only variants — use the LUT tab instead.

// Run a fit first to generate code

Pre-computed lookup table. Suitable for MCUs without FPU or when runtime computation budget is too tight for polynomial evaluation.

Run a fit first to generate the LUT.

// Run a fit first to generate LUT

LUT with linear interpolation between entries. Better accuracy than nearest- neighbour with only one extra multiply and subtract per call.

// Run a fit first to generate interpolation code