Closed mvidalgarcia closed 3 years ago
* Instead of hard-coding the "data.root" file name once more in `fitdata.inputs.data` clause, could we somehow say that `fitdata.inputs.data` is equal to `gendata.output`?
I'm not aware of a way to reference external rules outputs as you describe. The only solution that comes to my mind is keeping
"results/data.root"
as an input as we used to, so we can point to it likeconfig["data"]
. But I'm not sure, as this is not an actual workflow input.* Shall we add `.snakemake` to `.gitignore`?
No strong opinion. In the Snakefile we recommend using a separate folder
snakemake-local-run
to run the example locally to keep things isolated so I don't see an urgent need.
No strong opinion. In the Snakefile we recommend using a separate folder
snakemake-local-run
to run the example locally to keep things isolated so I don't see an urgent need.
It gets populated also when I run reana-dev run-example -w snakemake -c .
$ find .snakemake -type f
.snakemake/log/2021-08-16T145300.340515.snakemake.log
$ cat .snakemake/log/2021-08-16T145300.340515.snakemake.log
Job stats:
job count min threads max threads
------- ------- ------------- -------------
all 1 1 1
fitdata 1 1 1
gendata 1 1 1
total 3 1 1
that was my motivation....
But it's true that the directory is clean when one uses reana-client run
command, so regular usage should be OK.
closes reanahub/reana-workflow-engine-snakemake#5
fitdata
step didn't have"results/data.root"
as input so rule order was not respected.