I think there may be some issues with how submissions are scored and/or would appreciate an explanation/clarification of how the following submission is scored:
| year |
data_type |
source_index |
destination_index |
value |
| 20182019 |
depot_location |
0 |
|
|
| 20182019 |
refinery_location |
0 |
|
|
| 2018 |
biomass_forecast |
0 |
|
0 |
| 2019 |
biomass_forecast |
0 |
|
0 |
| 2018 |
biomass_demand_supply |
0 |
0 |
0 |
| 2019 |
biomass_demand_supply |
0 |
0 |
0 |
| 2018 |
pellet_demand_supply |
0 |
0 |
0 |
| 2019 |
pellet_demand_supply |
0 |
0 |
0 |
The above is essentially a null submission, with enough information filled in to avoid receiving an error code. This submission received a score of 80.79886 corresponding to a total cost of ~120007. No biomass is being transported and therefore the transportation costs should be 0 and the underutilization costs should be 120000 (one completely empty depot and one completely empty refinery). This then implies that the forecast error for this submission is 7. Note that as every site is either omitted or forecasted to produce no biomass the forecast cost in this case should equal the total biomass produced (assuming it is summed over all sites). This seems highly unlikely given that the total biomass for each year is approximately in the range of 250000-350000. This would suggest to me that either there is an error in the scoring function, or the forecast error is only summed over sites for which a forecast is provided?
*edited to proprely show submission
