vcm, vcx

Purpose

Computes an unbiased estimate a variance-covariance matrix.

Note

vcm() and vcx() have been replaced with functions varCovXS() and varCovMS() whose descriptions use more standard statistical nomenclature. vcx() and vcm() will continue to be available for backwards compatibility.

Format

vc = vcm(m)
vc = vcx(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 variance-covariance matrix) – unbiased estimate of the variance-covariance matrix.

Remarks

The variance-covariance matrix is computed as an unbiased estimator of the population variance-covariance. It is computed as the moment matrix of deviations about the mean divided by the number of observations minus one, \(N - 1\). For an observed variance-covariance matrix which uses \(N\) rather than \(N - 1\) see vcms() or vcxs().

Source

corr.src

See also

Functions momentd()