bpr_gibbs.matrix: Gibbs sampling for the BPR model using list x

In andreaskapou/mpgex: Methylation Profiles predict Gene Expression

Description

bpr_gibbs.matrix computes the posterior of the BPR model using auxiliary variable approach. Since we cannot compute the posterior analytically, a Gibbs sampling scheme is used.

Usage

## S3 method for class 'matrix'
bpr_gibbs(x, w_mle = NULL, basis = NULL,
  fit_feature = NULL, cpg_dens_feat = FALSE, w_0_mean = NULL,
  w_0_cov = NULL, gibbs_nsim = 20, gibbs_burn_in = 10, ...)

Arguments

x	An L x 3 matrix of observations, where 1st column contains the locations. The 2nd and 3rd columns contain the total trials and number of successes at the corresponding locations, repsectively.
w_mle	A vector of parameters (i.e. coefficients of the basis functions) containing the MLE estimates.
basis	A 'basis' object. See polynomial.object
fit_feature	Additional feature on how well the profile fits the methylation data.
cpg_dens_feat	Additional feature for the CpG density across the promoter region.
w_0_mean	The prior mean hyperparameter for w
w_0_cov	The prior covariance hyperparameter for w
gibbs_nsim	Optional argument giving the number of simulations of the Gibbs sampler.
gibbs_burn_in	Optional argument giving the burn in period of the Gibbs sampler.
...	Additional parameters

Value

A list containing the following elements:

• w_opt: Optimized values for the coefficient vector w. The length of the result is the same as the length of the vector w.

• basis: The basis object.

See Also

bpr_optim, bpr_optim.list

Examples

basis <- polynomial.object(M=2)
w <- c(0.1, 0.1, 0.1)
w_0_mean <- rep(0, length(w))
w_0_cov <- diag(10, length(w))
data <- bpr_data[[1]]
out_opt <- bpr_gibbs(x = data, w_mle = w, w_0_mean = w_0_mean,
                     w_0_cov = w_0_cov, basis = basis)

basis <- polynomial.object(M=0)
w <- c(0.1)
w_0_mean <- rep(0, length(w))
w_0_cov <- diag(10, length(w))
data <- bpr_data[[1]]
out_opt <- bpr_gibbs(x = data, w_mle = w, w_0_mean = w_0_mean,
                     w_0_cov = w_0_cov, basis = basis)

andreaskapou/mpgex documentation built on May 12, 2019, 3:33 a.m.