ENH: stats: transformations of random variables and more

Hi Team,

A series of PRs would add support for various transformations of random variables.

  • gh-21746 adds the exponential, logarithm, and reciprocal of a random variables via stats.exp, stats.log, and ContinuousDistribution.__rtruediv__.
  • gh-21777 adds the absolute value of a random variable by definining ContinuousDistribution.__abs__ and possibly adding stats.abs. (A simple follow-up could define __pow__ and __rpow__.)
  • gh-21752 implements truncated distributions as scipy.stats.TruncatedDistibution.
  • gh-21790 implements order statistic distributions as scipy.stats.OrderStatisticDistribution.

Also, gh-21871 would add support for generation of QMC samples from statistical distributions.

If you’re interested, please join the conversation in these PRs.

Thanks!
Matt

If no one objects, I plan to merge gh-21746 mentioned above today so we can get it in before we branch for 1.15.