ACE: Alternating Conditional Expectations

In Tsphere: Transposable Sphering for Large-Scale Inference with Correlated Data.

Description

Predicts values for scattered missing elements according to the matrix-variate normal distribution. Internal use.

Usage

ACE(x, sig, delt, sigi, delti, M, thr = 1e-04, maxit = 1000)

Arguments

x	Data matrix with scattered missing values. Missing values should be denoted with "NA".
sig	Row covariance matrix.
delt	Column covariance matrix.
sigi	Row precision matrix.
delti	Column precision matrix.
M	Mean matrix.
thr	Convergence threshold.
maxit	Maximum number of iterations.

Details

For internal use.

Value

x	Matrix of predicted values.
iter	Number of iterations until convergence.

Author(s)

Genevera I. Allen

References

G. I. Allen and R. Tibshirani, "Transposable regularized covariance models with an application to missing data imputation", Annals of Applied Statistics, 4:2, 764-790, 2010.

See Also

CVcov

Tsphere documentation built on May 2, 2019, 3:32 p.m.