You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
This library can make predictions about data using a technique called polynomial regression.
Polynomial regression uses a technique called Gaussian-Jordan elimination, which creates a predictive model that more accurately fits non-linear data.
How to use
Let's say you have your typical cartesian coordinates (x and y coordinates)
constdata=[{x : 5,y : 8},{x : 9,y : 12}// and so on...];
This library will read this data, and then make a prediction about a y value, given an x.
//This library is a UMD module (thanks webpack!)importPolynomialRegressionfrom"js-polynomial-regression";//Factory function - returns a PolynomialRegression instance. 2nd argument is the degree of the desired polynomial equation.constmodel=PolynomialRegression.read(data,3);//terms is a list of coefficients for a polynomial equation. We'll feed these to predict y so that we don't have to re-compute them for every prediction.constterms=model.getTerms();//10 is just an example of an x value, the second argument is the independent variable being predicted.constprediction=model.predictY(terms,10);
That's it! I've created an example using random data in the example folder of this repo. Please use the issues section to communicate any bugs, questions, or feature requests.
About
A javascript library that predicts the value of a dependent variable using polynomial regression (Gaussian-Jordan elimination)