In the course of writing up the project report from the summer school, i noticed that RoW processes were being over represented in market processes because they retained the global production volume after they'd been renamed from GLO to RoW.
I've added this fix (with its test) to my fork of Ocelot, which may (or may not) be worth pulling into the master.
As relabel_global_to_row renames transforming activities from GLO to RoW it now sums up the production volumes for the specific regions and subtracts them from the GLO/RoW production volume before adding it to the processed dataset
Coverage increased (+0.003%) to 85.01% when pulling a4eabb9c7626fa93021434c137777fa57873ff46 on pjamesjoyce:GLOtoRoW into fc7191459019fb3de1b1727dc4b17c31027ad230 on OcelotProject:master.
Coverage increased (+0.06%) to 85.072% when pulling 70d17e2111cd421c8311137322e50b1d47b773f1 on pjamesjoyce:GLOtoRoW into fc7191459019fb3de1b1727dc4b17c31027ad230 on OcelotProject:master.
Coverage increased (+0.003%) to 85.01% when pulling e4695c1380fff172f8b1b15665f24d964093b858 on pjamesjoyce:GLOtoRoW into fc7191459019fb3de1b1727dc4b17c31027ad230 on OcelotProject:master.
In the course of writing up the project report from the summer school, i noticed that RoW processes were being over represented in market processes because they retained the global production volume after they'd been renamed from GLO to RoW.
I've added this fix (with its test) to my fork of Ocelot, which may (or may not) be worth pulling into the master.
As
relabel_global_to_row
renames transforming activities from GLO to RoW it now sums up the production volumes for the specific regions and subtracts them from the GLO/RoW production volume before adding it to the processed dataset