pyab_experiment.utils.stats

Binomial CI estimates

Module Contents

Functions

confidence_interval(→ tuple[float, float])	Return the mean and the confidence interval of Bernoulli trials.
probit(→ float)	Compute the probit of a given value.

pyab_experiment.utils.stats.confidence_interval(n: int = 10, p: float = 0.5, confidence: float = 0.95, method: str = 'agresti-coull') → tuple[float, float]

Return the mean and the confidence interval of Bernoulli trials. See `https://en.wikipedia.org/wiki/Binomial_distribution#Confidence_intervals`_

pyab_experiment.utils.stats.probit(alpha: float = 0.5) → float

Compute the probit of a given value.

The probit function is the inverse of the cumulative distribution function (CDF) of the standard normal distribution. It maps a probability (between 0 and 1) to a corresponding value on the real number line.

Parameters:

alpha (float) – The probability value for which to compute the probit. Default is 0.5, which corresponds to the median of the standard normal distribution.

Returns:

The probit value corresponding to the input probability.

Return type:

float

Note

See `https://en.wikipedia.org/wiki/Probit for the approximation`_ of using the logit function.

Examples

>>> probit(0.5)
0.0
>>> probit(0.975)
1.959963984540054