Closed donaldcampbelljr closed 6 months ago
Some notes: -Looper reads the config file and sets arguments based on the config file: https://github.com/pepkit/looper/blob/6523e4b525570aa0f7b3013d19a049d1866a28df/looper/cli_looper.py#L593-L619
However, when Looper creates a Project, not all of these lines from the config file become attributes on the Project.
Specific example is with the addition of pipestat in the looper config file:
pep_config: ./project/project_config.yaml
output_dir: ./results
pipeline_interfaces:
sample: ./pipeline_pipestat/pipeline_interface.yaml
pipestat:
results_file_path: results.yaml
The project does not contain the results_file_path
. It does, however, contain a boolean pipestat_configured = True
.
Further in the process, SubmissionConductor builds the looper namespace: https://github.com/pepkit/looper/blob/6523e4b525570aa0f7b3013d19a049d1866a28df/looper/conductor.py#L462-L500
It appears that only select items are added to the Looper namespace. The config file that is added is the project_config
NOT the looper_config
.
Question: Is the desire to have everything in the config files (looper and project and pep configs) be an attribute of the looper namespace such that it could be accessed if desired?
I believe looper namespace is being built properly now. In the above example, I mentioned results_file_path
as an example but that Is actually for pipestat and is contained within the pipestat namespace.
Original comment: looper variable namespaces need to be updated; in particular, does the looper namespace get what's in the .looper config file now?