search

djgroen / FabFlee

FabSim3 plugin for automated Flee-based simulations
BSD 3-Clause "New" or "Revised" License
4 stars 8 forks source link

Config_files inconsistencies in validation folder #28

Open cspgdds opened 10 months ago

cspgdds commented 10 months ago

There are a few conflict instances with inconsistencies in config_files/validation folder

cspgdds commented 10 months ago
  1. mali2017_adm1 (1735 days) includes source_data for camps with a decreasing number of refugees over the simulation period, which are normalized using registration_corrections.csv @djgroen
cspgdds commented 10 months ago
  1. nigeria2016 (1885 days) contains no towns, only conflict_zones and camps. It might require a new routes.csv file, if towns are added to the conflict instance. @mzrghorbani
cspgdds commented 10 months ago
  1. nigeria2022_adm2 (395 days) contains no towns, only conflict_zones and camps. It might require a new routes.csv file, if towns are added to the conflict instance. @cspgdds
cspgdds commented 10 months ago
  1. syria2013 (1095 days) includes only one town. It might require a new routes.csv file, if additional towns are added to the conflict instance. @alirezajahani60
mzrghorbani commented 10 months ago
  1. nigeria2016 (1885 days) contains no towns, only conflict_zones and camps. It might require a new routes.csv file, if towns are added to the conflict instance. @mzrghorbani

To add towns, is it preferred to reconstruct the ACLED file? If yes, please specify the timeline for nigeria2016. @cspgdds

djgroen commented 10 months ago
  1. mali2017_adm1 (1735 days) includes source_data for camps with a decreasing number of refugees over the simulation period, which are normalized using registration_corrections.csv @djgroen

There are some other issues here as well, such as camps with starting populations. I'll see if I can resolve all this in a sensible manner.

alirezajahani60 commented 10 months ago
  1. syria2013 (1095 days) includes only one town. It might require a new routes.csv file, if additional towns are added to the conflict instance. @alirezajahani60

This model is in adm1 level, in which conflict_zones represents provinces. Do you think it is still required to add towns?