logit

scikitplot.cexperimental._cy_cexperimental.logit(x0)

Compute the logit (inverse sigmoid) of a scalar input.

The logit function transforms a probability in the interval (0, 1) into a real-valued number. It is defined as:

\[\operatorname{logit}(x) = \log\left(\frac{x}{1 - x}\right)\]

This function is commonly used in logistic regression and other probabilistic models.

Parameters:

x0dfg_number_t

Scalar input value in the range (0, 1). Typically a float, double, or long double.

Returns:

dfg_number_t

The logit of the input value, returned with the same type as x0.

See also

expit

The sigmoid (logistic) function.

log_expit

Logarithm of the sigmoid function.

Notes

The input must be strictly within the interval (0, 1) to avoid mathematical domain errors (e.g., division by zero or log of zero).

Added in version 0.3.9.

Examples

>>> logit(0.5)
0.0

>>> logit(0.8)
1.3862943611198906