Since the Snakemake_model_eval workflow is central to what clm will be used for, it should be picked up by default by snakemake. If only training/sampling is needed, then the workflow can be truncated on the command line (using --until .. or similar).
In the end, I feel like the number of steps needed to go from raw data to graphs should be minimal (preferably just one).
We can also look into structuring steps in the workflow, and following other best practices to get a handle on the increasing complexity of Snakefile.
Since the
Snakemake_model_evalworkflow is central to whatclmwill be used for, it should be picked up by default bysnakemake. If only training/sampling is needed, then the workflow can be truncated on the command line (using--until ..or similar).In the end, I feel like the number of steps needed to go from raw data to graphs should be minimal (preferably just one).
We can also look into structuring steps in the workflow, and following other best practices to get a handle on the increasing complexity of
Snakefile.