Open
Description
Below is a proposal.
README:
- 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:
- Contributor guide
- Directory structure
- Fetch Viash & Nextflow
- Build components
- Fetch test resources
- 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?