dash_smooth: Adaptive smoothing using Beta adaptive shrinkage

In kkdey/dashr: Dirichlet Adaptive Shrinkage (dash) of Compositional data with applications in smoothing

Description

Given a vector of counts, which are noisy estimates of an underlying Poisson counts data, the function performs adaptive smoothing of the counts by fitting a Beta adaptive shrinkage model.

Usage

dash_smooth(x, concentration = NULL, pi_init = NULL, reflect = TRUE,
  squarem_control = list(), dash_control = list(), progressbar = TRUE)

Arguments

x,	a vector of counts
concentration	a vector of concentration scales for different Dirichlet compositions. Defaults to NULL, in which case, we append concentration values of Inf, 100, 50, 20, 10, 5, 2, 1, 0.5 and 0.1.
pi_init	An initial starting value for the mixture proportions. Defaults to same proportion for all categories.
reflect	Boolean indicating if the vector is padded by a reflection of the tail or the tailmost value so that the padded vector has length a power of 2.
squarem_control	A list of control parameters for the SQUAREM/IP algorithm, default value is set to be control.default=list(K = 1, method=3, square=TRUE, step.min0=1, step.max0=1, mstep=4, kr=1, objfn.inc=1,tol=1.e-07, maxiter=5000, trace=FALSE).
dash_control	A list of control parameters for determining the concentrations and prior weights and fdr control parameters for dash fucntion.
progressbar	Boolean indicating whether to show the progress bar for the code run or not. Defaults to TRUE

Details

The input to dash-smooth is a vector of counts which are noisy versions of a smooth process. We fit a multiscale model on these counts and the message flow proportions are assumed to have a flexible mixture Beta prior centered around mean of 0.5.

Value

Returns a list of the following items estimate: The adaptively smoothed values of the counts vector x. pi_weights: The mixture proportions estimated from different levels of multiscale model. loglik: The loglikelihood value of the fitted model.

Examples

mu <- c(rep(10, 50), rep(20, 50), rep(30, 50), rep(10, 50))
x <- mu + rnorm(200, 0, 1)
out <- dash_smooth(x)
out$estimate

kkdey/dashr documentation built on May 3, 2019, 9:38 p.m.