Whenever there is a new release of https://github.com/pypsa/technology-data we should update the ariadne-data/costs/{mean,optimist,pessimist} files. Ideally that would happen automatically in the snakemake workflow, but creating the files requires executing the technology-data workflow with different configurations for the data uncertainty.
I just wasted an hour trying to find out why a new key was not present in the cost data, just to realize that retrieve_costs is disabled in pypsa-ariadne.
We should have at least a helpful error message that checks if the version in ariadne-data is up to date with the version specified in default-config.
To automate updating the ariadne-data/costs we could add a CI.
For each new release of technology-data, a runner is started, executes the technology data workflow with the required configs, and places the output in ariadne-data.
Technology-data is going to be a package. Once that's done I will write a rule to put together the cost data.
Until then, I manually upload the missing two cost horizons.
Whenever there is a new release of https://github.com/pypsa/technology-data we should update the ariadne-data/costs/{mean,optimist,pessimist} files. Ideally that would happen automatically in the snakemake workflow, but creating the files requires executing the technology-data workflow with different configurations for the data uncertainty.
I just wasted an hour trying to find out why a new key was not present in the cost data, just to realize that retrieve_costs is disabled in pypsa-ariadne.
We should have at least a helpful error message that checks if the version in
ariadne-data
is up to date with the version specified in default-config.To automate updating the
ariadne-data/costs
we could add a CI.For each new release of technology-data, a runner is started, executes the technology data workflow with the required configs, and places the output in
ariadne-data
.Maybe this is something @lkstrp could tackle?