ss.phi: SS.phi

In mpoll/scale: The Scaleable Langevin Exact Algorithm -- Reference Implementation

Usage

ss.phi(eta.pos, dapts, ss.size)

Arguments

eta.pos
dapts
ss.size

Examples

##---- Should be DIRECTLY executable !! ----
##-- ==>  Define data, use random,
##--	or do  help(data=index)  for the standard data sets.

## The function is currently defined as
function (eta.pos, dapts, ss.size) 
{
    beta.pos <- un.scale.transform(eta.pos)
    factor <- dsz/ss.size
    data1 <- data.eval(beta.pos, dapts[1, , "dapt.1"], factor = factor)
    data1.star <- data.eval(beta.star, dapts[1, , "dapt.1"], 
        factor = factor)
    data2 <- data.eval(beta.pos, dapts[1, , "dapt.2"], factor = factor)
    data2.star <- data.eval(beta.star, dapts[1, , "dapt.2"], 
        factor = factor)
    data3 <- data.eval(beta.pos, dapts[1, , "dapt.3"], factor = factor)
    data3.star <- data.eval(beta.star, dapts[1, , "dapt.3"], 
        factor = factor)
    ((data1$grad.log.pi - data1.star$grad.log.pi) %*% t(2 * alpha.cent + 
        data2$grad.log.pi - data2.star$grad.log.pi) + sum(data3$lap.log.pi) - 
        sum(data3.star$lap.log.pi) + alpha.cent.sq + alpha.p.cent)/2
  }

mpoll/scale documentation built on Dec. 9, 2019, 7:15 a.m.