Improve the statistics on the product page (tracker)

[aleene]

Description

The statistics module now presents some standard information (mean, deviation, min, max, 10%, median). This could be extended to include parameters to detect/identify outliers. This can be used to identify products in the distribution with strange values.
Estimates for a beta distribution, which will provide better means and deviations as well for the nutrients. Nutrients lie on a bounded interval, which means that a normal distribution is not applicable.
It should be possible to retrieve the statistics data through an API.

Acceptance criteria

For a demo category we could define the acceptable values that can be checked;

What would a demo look like

The statistics information on a facet page could include the new information;

Notes

• Do not use upper limits in the statistics calculation;
• Suppress the 10% quantiles and replace them with 25% quartiles;

Tasks

• Define calculations for the required statistics
• Setup the calculations
• Expose the values on web
• Expose the values on API
• Use boundaries to raise product flags
• Add flags to data quality monitoring