pipestat output schema seems to lack path templates

[nsheff]

I'm reading about output schemas here:

https://pep.databio.org/looper/developer-tutorial/pipeline-interface-specification/#output_schema

This documentation seems to lack for me one of the main reasons of using the output schema: to automatically add variables to a sample object, populated via a template.

I was expecting to say somethign like

      properties:
        fragment_file:
          type: string
          path_template: "outs/{sample_name}_fragments.tsv.gz"

or something. What happened to this functionality?

[donaldcampbelljr]

I'm confused. The output schema defines the attributes for reported results via pipestat, not adding variables to a sample object in Looper.

[nsheff]

Those two concepts are very similar, I guess.

a reported result is basically an attribute of a sample...

I thought in the past I used these as part of a sample object. or at least, that's what I wanted to do when I wrote this.

[donaldcampbelljr]

Historically, Looper output schema used to define (via a template) where the paths would be resolved, like the above example.

Today, Pipestat has a results schemas which define what types the paths should be. It does not resolve variables within the fields per sample.

Desires:

1. create pep
2. run pipeline
3. load up pep again to have all the results resolved

But is there a way to this just in peppy where you load the PEP and it gets these results?

Right now a pipeline would load pipestat manager and its corresponding results.yaml to send items downstream for analysis.