This is a half-finished and somewhat lame visualization of a model for the upper bound of a uniform likelihood, with a Pareto prior. You can specify hyperparameters and visualize the resulting prior distribution, then you specify the true value of Θ and the number of samples to draw. In response, the posterior distribution is plotted.

This was mainly a way to play around with JavaScript and statistics, and there are some notable problems (sensible ranges, insensible inputs). I may revisit this little experiment later.

You can think of this as the "maximum", with the caveat that there's no probability mass below x_{m}. So it's more of a "minimum maximum". x_{m} > 0.

This is called the "tail index" and controls the thickness of the tail. The larger it gets, the thinner the tail. You can think of it as the number of samples. α > 0.

The upper bound on the uniform distribution.

The number of samples to take. (It's possible to set this too high, actually.)