vector.calculategeneralleastsquares.md

Home > @josh-brown/vector > calculateGeneralLeastSquares

calculateGeneralLeastSquares() function

Calculates a regression model for an arbitrary function.

Signature:

export declare function calculateGeneralLeastSquares<S>(dataPoints: Vector<S>[], functionTemplate: ApproximationFunctionTemplate<S>, numberOfTerms: number): LeastSquaresApproximation<S>;

Parameters

Parameter	Type	Description
dataPoints	Vector<S>[]	The data used to construct the approximation
functionTemplate	ApproximationFunctionTemplate<S>	A higher-order function which takes a vector of coefficients and yields a new function which takes a vector of independent variables to produce a value for the dependent variable
numberOfTerms	number	The number of coefficients needed to produce the approximation function

Returns:

LeastSquaresApproximation<S>

• the result of the linear regression

Remarks

The result is on object which has: - coefficients: a vector whose entries correspond to the coefficients which must be plugged into the function template to yield the best approximation function - approximationFunction: a function which takes a vector of the independent variable values, and returns the predicted value of the dependent variable