ENO: Normal distribution with empirical mean and variance

In davharris/mistnet2: Neural Networks with Latent Random Variables

Description

This distribution produces a Gaussian distribution whose moments are dynamically set to match mean(x) and sd(x). It is used for empirical-Bayesian descriptions of the distribution of x (similar to the distribution of coefficients in a mixed effects model).

Usage

ENO(by)

dENO(x, log, by, ...)

rENO(n, x, ...)

Arguments

by	one of "row", "col", or "mat". Should the empirical moments be estimated for each row, each column, or for the full matrix?
x	vector of quantiles
log	logical; if TRUE, probabilities p are instead given as log(p)
...	additional arguments passed to or from other methods (currently not used)
n	number of samples

davharris/mistnet2 documentation built on May 14, 2019, 9:28 p.m.