[ ] Point users to the user guidelines on the website
[ ] Point contributors to the contributor guidelines
[ ] Add a set of commands to quickly get started with modifying the pipeline, e.g.:
# download viash and nextflow
bin/init
# download test resources
bin/viash run src/download/sync_test_resources/config.vsh.yaml
# run unit test for a certain component
bin/viash_test --query pca
# run a component on some data
bin/viash run src/dimred/pca/config.vsh.yaml -- --input resources_test/pbmc_1k_v3_protein/mms.h5mu --output foo.h5mu
[ ] Run a pipeline from the CLI
[ ] Run a pipeline from nf-tower
Website:
[x] Contributor guide
[x] Directory structure
[x] Fetch Viash & Nextflow
[x] Build components
[x] Fetch test resources
[x] Run tests
[ ] Create new components (refer to viash.io)
[ ] Create new pipelines (refer to viash.io)
[ ] User guide (separate page, should not be in the contributors guide)
[ ] Run pipeline from CLI
[ ] Run pipeline from NF tower
[ ] Ingesting large param sets with param_list
[ ] To configure infrastructure related settings, create a profile with additional labels:
[ ] Setting resource constraints with labels (e.g. highmem, highcpu)
[ ] Singularity, Docker, ...?
[ ] Run on HPC cluster, cloud, ...
[ ] Deploying pipelines on premises -- get in touch?
Below is a proposal.
README:
Website:
param_list