cv.gpca: CV for generalized PCA

In andland/generalizedPCA: Generalized PCA

Description

Run cross validation on dimension and M for generalized PCA

Usage

cv.gpca(x, ks, Ms = seq(2, 10, by = 2), family = c("gaussian", "binomial",
  "poisson", "multinomial"), weights, folds = 5, quiet = TRUE, ...)

Arguments

x	matrix of either binary, count, or continuous data
ks	the different dimensions k to try
Ms	the different approximations to the saturated model M to try
family	exponential family distribution of data
weights	an optional matrix of the same size as the x with data weights
folds	if folds is a scalar, then it is the number of folds. If it is a vector, it should be the same length as the number of rows in x
quiet	logical; whether the function should display progress
...	Additional arguments passed to generalizedPCA

Value

A matrix of the CV log likelihood with k in rows and M in columns

Examples

# construct a low rank matrix in the natural parameter space
rows = 100
cols = 10
set.seed(1)
mat_np = outer(rnorm(rows), rnorm(cols))

# generate a count matrix
mat = matrix(rpois(rows * cols, c(exp(mat_np))), rows, cols)

## Not run: 
loglikes = cv.gpca(mat, ks = 1:9, Ms = 3:6, family = "poisson", quiet = FALSE)
plot(loglikes)

## End(Not run)

andland/generalizedPCA documentation built on May 12, 2019, 2:42 a.m.