Bug or not: BEST Notebook: Difference of means of posterior probability in plot vs. text

,

Dear Community,

I believe there is a bug/typo in PyMC’s BEST Notebook. Specifically, when describing the difference of means of the posterior probability in the text that accompanies the posterior plot, the author states that, for the difference in means, 99.4% of the posterior probability is greater than zero. However, the plot just below shows this number as 98.4%.

Am I not understanding something or should the text actually say “98.4%”?

If this indeed is a bug, I am willing to fix it, but I would love to have your feedback first. Thanks a lot!

Florian

To be safe we should say larger than 97% as the value is an stochastic estimation. Feel free to send a pull request.

Thanks @junpenglao, that is a great suggestion. I will change the notebook accordingly and open a pull request.

On a second thought though, is there a possibility to seed PyMC3, so that results of this demo notebook are reproducible? I would imagine seeding PyMC3 on purpose to be a welcome characteristic of demo notebooks for folks trying to reproduce them on their own machines.

This is not a bug.

The explanation is given in the (annotated) paragraph:

For these comparisons, it is useful to use zero as a reference value (ref_val); providing this reference value [i.e. 0] yields cumulative probabilities [as in cdf] for the posterior distribution on either side of the value [i.e. vertical line at ref_val=0]. Thus, for the difference in means, 99.4% of the posterior probability is greater than zero…

In other words, 99.4% == values to the right of vertical line - values to the left.