mixash: Main Adaptive SHrinkage function
In mengyin/mixash: What the package does (short line)

Description Usage Arguments Details Value Examples

View source: R/mixash.R

Takes vectors of estimates (betahat) and their standard errors (sebetahat), and applies shrinkage to them, using Empirical Bayes methods, to compute shrunk estimates for beta.

mixash(betahat, sebetahat, df, pilik, method = c("shrink", "fdr"),
  mixcompdist = c("normal", "uniform", "halfuniform"), lambda1 = 1,
  lambda2 = 0, nullcheck = FALSE, randomstart = FALSE, pointmass = TRUE,
  onlylogLR = FALSE, singlecomp = FALSE, SGD = TRUE,
  prior = c("uniform", "nullbiased"), mixsd = NULL, gridmult = sqrt(2),
  minimaloutput = FALSE, g = NULL, control = list())

`betahat,`	a p vector of estimates
`sebetahat,`	a p vector of corresponding standard errors
`control`	A list of control parameters for the SQUAREM 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). User may supply changes to this list of parameter, say, control=list(maxiter=10000,trace=TRUE)
`method:`	specifies how ash is to be run. Can be "shrinkage" (if main aim is shrinkage) or "fdr" (if main aim is to assess fdr or fsr) This is simply a convenient way to specify certain combinations of parameters: "shrinkage" sets pointmass=FALSE and prior="uniform"; "fdr" sets pointmass=TRUE and prior="nullbiased".
`mixcompdist:`	distribution of components in mixture ("normal", "uniform" or "halfuniform")
`lambda1:`	multiplicative "inflation factor" for standard errors (like Genomic Control)
`lambda2:`	additive "inflation factor" for standard errors (like Genomic Control)
`nullcheck:`	whether to check that any fitted model exceeds the "null" likelihood in which all weight is on the first component
`df:`	appropriate degrees of freedom for (t) distribution of betahat/sebetahat
`randomstart:`	bool, indicating whether to initialize EM randomly. If FALSE, then initializes to prior mean (for EM algorithm) or prior (for VBEM)
`pointmass:`	bool, indicating whether to use a point mass at zero as one of components for a mixture distribution
`onlylogLR:`	bool, indicating whether to use this function to get logLR. Skip posterior prob, posterior mean, lfdr...
`singlecomp:`	bool, indicating whether to use a single inverse-gamma distribution as the prior distribution of the variances
`SGD:`	bool, indicating whether to use the stochastic gradient descent method to fit the prior distribution of the variances
`unimodal:`	unimodal constraint for the prior distribution of the variances ("variance") or the precisions ("precision")
`prior:`	string, or numeric vector indicating Dirichlet prior on mixture proportions (defaults to "uniform", or 1,1...,1; also can be "nullbiased" 1,1/k-1,...,1/k-1 to put more weight on first component)
`mixsd:`	vector of sds for underlying mixture components
`gridmult:`	the multiplier by which the default grid values for mixsd differ by one another. (Smaller values produce finer grids)
`minimal_output:`	if TRUE, just outputs the fitted g and the lfsr (useful for very big data sets where memory is an issue)
`g:`	the prior distribution for beta (usually estimated from the data; this is used primarily in simulated data to do computations with the "true" g)

See readme for more details

a list with elements fitted.g is fitted mixture logLR : logP(D|mle(pi)) - logP(D|null)

beta = c(rep(0,100),rnorm(100))
sebetahat = abs(rnorm(200,0,1))
betahat = rnorm(200,beta,sebetahat)
beta.ash = ash(betahat, sebetahat)
summary(beta.ash)
plot(betahat,beta.ash$PosteriorMean,xlim=c(-4,4),ylim=c(-4,4))