LogUniform 1.0.1

Implementation of the log-uniform and modified log-uniform distributions

A simple implementation of the log-uniform and modified log-uniform distributions.

How to use

Install it from pip (LogUniform only depends on numpy)

pip install LogUniform

and it’s ready to use from Python

import loguniform

LogUniform comes with two simple classes, LogUniform and ModifiedLogUniform. They are intended to mimic the API of scipy.stats (actually, the log-uniform distribution is already implemented in scipy.stats.reciprocal; the two implementations are compatible).

from loguniform import LogUniform, ModifiedLogUniform

d1 = LogUniform(a=1, b=1000)
d2 = ModifiedLogUniform(knee=1, b=1000)

both distributions d1 and d2 now have methods

• pdf(x) and logpdf(x)

• cdf(x)

• rvs(size)

to calculate the probability density function (and its logarithm), the cumulative density function, and to get random samples from the distribution.

License

Copyright 2018 João Faria.

LogUniform is free software made available under the MIT License. For details see the LICENSE file.