mexpit: mexpit - Convert log-scale numbers to probabilities

In rxode2: Facilities for Simulating from ODE-Based Models

mexpit – Convert log-scale numbers to probabilities

Description

This function converts log-scale numbers to probabilities.

Usage

mexpit(...)

dmexpit(...)

Arguments

...	numeric log-scale numbers to convert to probabilities.

Details

The probabilities are calculated using the following equation:

p_i = \frac{e^{x_i}}{1+\sum_{j=1}^{N-1} e^{x_j}}

This ensures one remaining probability will add to one, that is

p_N = \frac{1}{1+\sum_{j=1}^{N-1} e^{x_j}}

For the function dmexpit(), the element-wise derivatives are calculated; that is, it returns the diagonal of the Jacobian matrix, d p_i / d x_i, not the full Jacobian with off-diagonal terms.

Value

Probabilities that add up to a number less than 1.

Author(s)

Matthew L. Fidler

Examples

m <- mlogit(0.1, 0.2, 0.3)
mexpit(m)

# derivatives
dmexpit(m)

p <- mexpit(-3, 0.5, 3)
mlogit(p)

rxode2 documentation built on July 28, 2026, 5:08 p.m.