Open jkikstra opened 6 months ago
Things done for MESSAGE-pretending-REMIND file:
scenarios
folder: REMIND-MAgPIE-CEDS-RESCUE-Tier1-2024-04-25.csv
, REMIND-MAgPIE-CEDS-RESCUE-Tier1-2023-12-13.csv
(input emissions scenario files), regionmappingH12.csv
(model region mapping)scenarios
folder: harmonization_overrides.xlsx
(to only column names, empty rowsscenarios
folder: pretend_REMIND_is_MESSAGE.csv
(input emissions scenario file) and region_definitions_message.csv
(model region mapping)config.yml
fields:
pip install -e .[tests]
in a new conda environment,pytest tests
yields:`mamba env create -f environment.yml
/mamba env create -f environment.yml
cd notebooks
, then runjupytext --sync workflow.py
(NB delete comment) --> worksdownscaled = workflow.harmonize_and_downscale()
) works (may take a bit due to downloading and loading the first time - 13 mins on jarmo's laptop and IIASA wifi)AttributeError: 'DataFrame' object has no attribute 'include_in_total'
- because we expect "variabledefs-region" instead of "variabledefs-rescue"KeyError: 'global'
in when runningworkflow.harmonize_and_downscale()
-- new commits came in on the main branch -- after updating the environment using
mamba env update -f environment.yml
we continue-- now start changing things for
pretend_REMIND_is_MESSAGE.csv
--ur = set_openscm_registry_as_default()
, we getNo module named 'openscm_units'
, so quick-fix ismamba install -c conda-forge openscm-units
+restart kernel --> real fix is by adding openscm-units to the environment.yml laterindexraster_region
, we get the errorMapping undefined for some
countrys: bes, cuw, ggy, jey, sxm
--> so it is crucial to list somewhere all the countries that need to be defined