Logit

The logit function, an abbreviation of logistic unit, is the quantile function of the standard logistic distribution and the inverse of the standard logistic function. It is defined as logit(p) = ln(p over (1 minus p)) for a probability p between 0 and 1, which is why it is also called the log-odds function, mapping probabilities onto the full range of real numbers. The logit serves as the canonical link function for the Bernoulli distribution in generalized linear models, most notably in logistic regression, and it is central to the probabilistic Rasch model used in psychological and educational measurement.