Regression Analysis

Fitting models to data points using SciMathJS.

API Reference

linearRegression

Performs a simple linear regression using Ordinary Least Squares.

Result Object:

• slope: number
• intercept: number
• rSquared: number
• free(): void (WASM cleanup)

Signature:

function linearRegression(x: Float64Array | number[], y: Float64Array | number[]): LinearRegressionResult

---

fitLinear

Convenience method that returns results as a simple array.

Signature:

function fitLinear(x: Float64Array | number[], y: Float64Array | number[]): [slope: number, intercept: number, rSquared: number]

---

fitPolynomial

Fits a polynomial of specified order to data points.

Signature:

function fitPolynomial(x: Float64Array | number[], y: Float64Array | number[], order: number): Float64Array | null

---

fitGaussians

Fits a sum of $N$ Gaussian curves to the data using the Levenberg-Marquardt algorithm.

Signature:

function fitGaussians(
  x: Float64Array | number[], 
  y: Float64Array | number[], 
  initial: number[]
): Float64Array

• initial: Array of parameters sorted as [amp1, mu1, sigma1, amp2, mu2, sigma2, ...].
• Returns: Optimized parameters in the same format.

---

fitExponential / fitLogarithmic

Non-linear regression models for specific data patterns.

Signatures:

function fitExponential(x: Float64Array | number[], y: Float64Array | number[]): [a: number, b: number] | null
function fitLogarithmic(x: Float64Array | number[], y: Float64Array | number[]): [a: number, b: number] | null