map_pi_rho_nonnull_qge: Estimate pi and mixture proportions under a specified model

In jean997/cause: CAUSE: Causal Analysis Using Summary Effect Estimates

Description

Estimare the MAP for rho and the mixing proportions using coordinate descent. Causal effect (gamma) and the effect of U (eta) are fixed at zero.

Usage

map_pi_rho_nonnull_qge(
  X,
  mix_grid,
  sigma_g,
  qalpha,
  qbeta,
  q_start = qbeta(0.5, qalpha, qbeta),
  gamma_start = 0,
  eta_start = 0,
  rho_start = 0,
  tol = 1e-05,
  n.iter = 20,
  null_wt = 10,
  z_prior_func = function(z) {     dnorm(z, 0, 0.5, log = TRUE) },
  q_prior_func = function(q) {     dbeta(q, qalpha, qbeta, log = TRUE) },
  gammaeta_prior_func = function(g) {     dnorm(g, 0, sigma_g, log = TRUE) },
  optmethod = c("mixSQP", "mixIP")
)

Arguments

X	An object of class cause_data containing data for the two traits.
mix_grid	An object of class cause_grid containing variance pair candidates
rho_start	Starting value for rho
null_wt	Specifies the prior weight on the first entry of grid
z_prior_func	Prior function for z = arctanh(rho)

jean997/cause documentation built on Dec. 25, 2021, 10 p.m.