Overview > klyn.math > NDArray > covariance
Version 0.2.1 Overview Index Tree Deprecated
klyn.math.NDArray.covariance
method
public covariance(rowVariables as Boolean = true, ddof as Long = 1l, ignoreNaN as Boolean = false) as NDArray<Double>
Description

Covariance matrix of 1D/2D observations, with O(variables squared) output. Columns are not materialized. NaN removal, if requested, is pairwise; pairwise deletion does not guarantee a positive-semidefinite matrix.

Parameters
ParameterDescription
rowVariablesTrue: rows are variables, columns observations; false reverses them.
ddofNon-negative correction; default 1 for sample covariance.
ignoreNaNDrop NaN pairs; otherwise propagate NaN.
Returns

Square Double matrix; one variable produces a [1,1] matrix.

Throws
  • ValueException for unsupported rank or insufficient observations.
Example

matrix = observations.covariance(rowVariables=false)