For \(x = logit(p)\), this function returns \(log(p)\) and \(log(1-p)\). Special care is taken to ensure accuracy when coordinates are numerically close to \(0\) or \(1\).
Usage
LogPq(x)
Arguments
x
\(Logit(p)\) where \(p\) is a vector of probabilites
# NOT RUN {#Find log(p) and log(q) for x = logit(0.2)a <- log(0.2/(1 - 0.2))
LogPq(x = a)
#Find log(p) and log(q) for x = logit(1e-4)b <- log(1e-4/(1 - 1e-4))
LogPq(x = b)
# }