RegModels.MulLinEval Method

Overload List

#	Signature	Description
1	procedure MulLinEval(const B: TDoubleArray; const X: TMtx; const YHat: TVec; Constant: Boolean);	Evaluates b[0] + b[1]x[0] + b[2]x[1] + ... for given values in X vector, parameters in B array, and returns the results in XHat vector.
2	function MulLinEval(const B: TDoubleArray; const X: TVec; Constant: Boolean): Double;	Evaluates multiple linear function.

Overload 1: procedure MulLinEval(const B: TDoubleArray; const X: TMtx; const YHat: TVec; Constant: Boolean);

Evaluates b[0] + b[1]*x[0] + b[2]*x[1] + ... for given values in X vector, parameters in B array, and returns the results in XHat vector.

#	Name	Type
1	B	TDoubleArray
2	X	TMtx
3	YHat	TVec
4	Constant	Boolean

Result: stored in self (calling object)

Remarks:

Size and Complex properties of XHat vector are adjusted automatically.

Note
Use this version if you want to evaluate multiple linear function for multiple values at the same time. This is a lot faster than calling single value version for each x value.

See Also: RegModels.MulLinFit

Overload 2: function MulLinEval(const B: TDoubleArray; const X: TVec; Constant: Boolean): Double;

Evaluates multiple linear function.

#	Name	Type
1	B	TDoubleArray
2	X	TVec
3	Constant	Boolean

Returns: Double - b[0] + b[1]*x[0] + b[2]*x[1] + ... for given value X and parameters in B array.

Examples

Uses MtxExpr, RegModels;
procedure Example;
var YHat: Vector;
X: Matrix;
begin
    X.SetIt(3,3,false,[2, -3, 5,
    1, 6, -4,
    8, 7, 9]);
    MulLinEval([1, 0.5, -2],X,YHat,false); // no constant term !
    // YHat = (-9.5, 12, -6.5)
end;