ohenery (version 0.1.1)

inv_smax: The inverse softmax function.

Description

The inverse softmax function: take a logarithm and center.

Usage

inv_smax(mu, g = NULL)

Arguments

mu

a vector of the probablities. Must be the same length as g if g is given. If mu and eta are both given, we ignore eta and use mu.

g

a vector giving the group indices. If NULL, then we assume only one group is in consideration.

Value

the centered log probabilities.

Details

This is the inverse of the softmax function. Given vector \(\mu\) for a single group, finds vector \(\eta\) such that $$\eta_i = \log{\mu_i} + c,$$ where \(c\) is chosen such that the \(\eta\) sum to zero: $$c = \frac{-1}{n} \sum_i \log{\mu_i}.$$

See Also

smax

Examples

# NOT RUN {
# we can deal with large values:
set.seed(2345)
eta <- rnorm(12,sd=1000)
mu <- smax(eta)
eta0 <- inv_smax(mu)
# }