regression-js

Regression-js is a javascript module containing a collection of linear least-squares fitting methods for simple data analysis. It can be found on github.

Installation

This module works on node and in the browser. It is available as the regression package on npm. It is also available on a CDN.

npm

npm install --save regression

Usage

  import regression from 'regression';
  const result regression.linear([[0, 1], [32, 67], [12, 79]]);
  const gradient = result.equation[0];
  const yIntercept = result.equation[1];
      

Data is passed into the model as an array. A second parameter can be used to configure the model. The configuration parameter is optional. `null` values are ignored. The precision option will set the number of significant figures the output is rounded to.

Configuration options

Below are the default values for the configuration parameter.

{
  order: 2,
  precision: 2,
}
      

Properties

• equation an array containing the coefficients of the equation

• string A string representation of the equation

• points an array containing the predicted data in the domain of the input

• r2 the coefficient of determination (R²)

• predict(x) This function will return the predicted value

API

Linear

equation: [gradient, y-intercept] in the form y = mx + c

Exponential

equation: [a, b] in the form y = ae^bx

Logarithmic

equation: [a, b] in the form y = a + b ln x

Power law

equation: [a, b] in the form y = ax^b

Polynomial

equation: [ai, .... , a0] in the form a_ix^j ... + a₀x⁰

        var data = [[0,1],[32, 67] .... [12, 79]];
        var result = regression.polynomial(data, { order: 3 });