Stats fixes 2018

As part of redevelopment of the Fixometer in 2018, we also discovered and and resolved a number of long-standing and subtle bugs in the calculation of CO2 and ewaste figures, which mean figures from Misc devices were being underreported.

For global figures, the weight and CO2 estimate were missing figures for Misc devices. And also in a small number of cases where a device has somehow been given an estimated weight and then changed to a non-Misc category, it was being included in the calculations twice.

For group specific pages within the Fixometer, the same issue was occuring with the group's stats shown in the key stats section.

For group statistics provided in iframes, the stats were being calculated in a different way, with a different bug. The emission ratio was being calculated wrongly. Misc devices were being included, but with a subtle bug resulting in the displacement ratio being applied twice.

There is also a bug specific to the group stats iframe that was used on the therestartproject homepage, and later enabled for groups to embed themselves, where any individual events that achieved over 1,000 kg of CO2 diverted was getting truncated to the figure before the ,. e.g. 1,600 kg becomes 1 kg, 2,000 becomes 2 kg etc. This has only effected a couple of groups.

In all cases, the calculation for average CO2 per kilo of misc device was slightly out. It should be calculated only from fixed non-misc devices.

It's worth noting that the calculations will always change slightly over time, as the emissions ratio is dynamically calculated based on number on non-misc fixed devices.