Inverse logistic function: Difference between revisions

From Calculus
No edit summary
No edit summary
 
Line 7: Line 7:
The function may be extended to a function <math>[0,1] \to [-\infty,\infty]</math> with the value at 0 defined as <math>-\infty</math> and the value at 1 defined as <math>\infty</math>.
The function may be extended to a function <math>[0,1] \to [-\infty,\infty]</math> with the value at 0 defined as <math>-\infty</math> and the value at 1 defined as <math>\infty</math>.


===Probabilistic interpretation====
===Probabilistic interpretation===


Given a probability <math>p</math> (strictly between 0 and 1) the inverse logistic function computes the logarithm of the corresponding odds. Explicitly, the odds corresponding to probability <math>p</math> are:
Given a probability <math>p</math> (strictly between 0 and 1) the inverse logistic function computes the logarithm of the corresponding odds. Explicitly, the odds corresponding to probability <math>p</math> are:

Latest revision as of 15:57, 31 May 2014

Definition

The inverse logistic function or log-odds function is a function from the open interval (0,1) to all of R defined as follows:

xln(x1x)

The function may be extended to a function [0,1][,] with the value at 0 defined as and the value at 1 defined as .

Probabilistic interpretation

Given a probability p (strictly between 0 and 1) the inverse logistic function computes the logarithm of the corresponding odds. Explicitly, the odds corresponding to probability p are:

p1p

The logarithm of the odds is therefore:

ln(p1p)

Key data

Item Value
default domain open interval (0,1)
range all of R
inverse function logistic function x11+ex