NodeJS-OLS

A linear algebra implementation of ordinary least squares multiple-variate regression as a Node JS module with optional controls for heteroskedasticity

Note: Although I have tested this for accuracy, please do not base any work on it without first testing it yourself.

Dependencies

NodeJS: http://nodejs.org/ Sylvester: https://github.com/NaturalNode/node-sylvester

Implementation

Just make sure you have NodeJS installed.

var ols = require('./ols.js'),
    sylvester = require('sylvester');

Your dependent variable must be in a matrix of the dimensions N x 1 where N is the number of observations, and your independent variables must be in a matrix of the dimensions N x K where K is your number of parameters.

var Y = $M([[1],[2],[3]]),
    X = $M([[1,2,3],[1,2,3],[1,2,3]]);

After that point, you can perform an OLS regression in the following way:

ols.reg(Y,X)

The third optional parameter allows for heteroskedasticity robust standard errors, T-statistics, and P-values. It is called in the following way:

ols.reg(Y,X,true)

The regression results are returned in an object with one key for the test overall, and one key for each parameter including a constant. The function will return an error if there is multicollinearity or if you attempt to estimate more parameters than there are observations.