Statisticsvar Method (InArrayDouble, InArrayDouble, Boolean, Int32)

public static RetArray<double> var(
	InArray<double> A,
	InArray<double> Weights = null,
	bool biased = false,
	int dim = -1
)

Public Shared Function var ( 
	A As InArray(Of Double),
	Optional Weights As InArray(Of Double) = Nothing,
	Optional biased As Boolean = false,
	Optional dim As Integer = -1
) As RetArray(Of Double)

Dim A As InArray(Of Double)
Dim Weights As InArray(Of Double)
Dim biased As Boolean
Dim dim As Integer
Dim returnValue As RetArray(Of Double)

returnValue = Statistics.var(A, Weights, 
	biased, dim)

Parameters

A

Type: ILNumericsInArrayDouble
Input array A.

Weights (Optional)

Type: ILNumericsInArrayDouble
[Optional] Vector of scaling factors, same length as working dimension of A, default: no scaling.

biased (Optional)

Type: SystemBoolean
[Optional] true: apply biased normalization to result, default: false (non-biased).

dim (Optional)

Type: SystemInt32
[Optional] Index of the dimension to operate along. If omitted operates along the first non singleton dimension (i.e. != 1).

Return Value

On scalar A a scalar 0 of the same shape as A is returned.

On empty A an empty array is returned, having the dimension to operate along reduced to length 1.

The parameters Weights, biased and dim are optional. Ommiting either one will choose its respective default value.

The result for biased = true is computed by the following formula:

r = (A - mean(A)); 
            var = sum(r * r) / A.D[dim];

r = (A - mean(A)); 
            var = sum(r * r) / (A.D[dim] - 1);

If Weights is given, the normalization is applied to r as follows:

w = w / sum(w); 
            r = A - sum(w * A);
            var = sum(w * (r * r));

[ILNumerics Statistics Toolbox]

Reference