vcms, vcxs#

Purpose#

Computes the observed variance-covariance matrix.

Note

vcms() and vcxs() have been replaced with functions varCovX() and varCovM() whose descriptions use more standard statistical nomenclature. vcxs() and vcms() will continue to be available for backwards compatibility.

Format#

vc = vcms(m)#

vc = vcxs(x)#

Parameters:

m (KxK moment (\(x'x\)) matrix) – A constant term MUST have been the first variable when the moment matrix was computed.
x (NxK matrix) – data

Returns:

vc (KxK matrix) – the observed-covariance matrix of m or x.

Remarks#

The variance covariance matrix is that of the input data matrix. It is computed as the moment matrix of deviations about the mean divided by the number of observations \(N\). For an unbiased estimator covariance matrix which uses \(N - 1\) rather than \(N\) see vcm() or vcx().

Examples#

// Compute observed variance-covariance matrix from data
x = rndn(100, 3);
vc = vcxs(x);
print vc;

// Compute from a moment matrix (constant term must be first column)
x = rndn(100, 3);
m = (ones(100, 1) ~ x)'(ones(100, 1) ~ x);
vc = vcms(m);
print vc;

Source#

corrs.src