[erlang-questions] PropEr testing of a fixed set of atoms

[Jesper Louis Andersen]

On Sun, Jun 1, 2014 at 1:03 AM, Jay Nelson <jay@REDACTED> wrote:

> I currently do the following to test it, assuming 6 tries will oversample
> the set of
> three values and nearly always test all three cases:
>

The key is to use collect/aggregate/classify. You want to print the
distribution of your test cases so you have confidence you got all variants
covered. If the distributions are skewed, you want to tune your generators
so they generate more nasty cases.

Coincidentally, this is why automatic derivation of generators isn't that
useful in practice.


-- 
J.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://erlang.org/pipermail/erlang-questions/attachments/20140604/4f44e417/attachment.htm>