Create automated work-flows

[tbhallett]

I propose the following two work flows to be developed using the comment-trigger workflow:

/scale-run

Run the model at scale (50k) using scale_run.py, 2010-2030, cycling through:

• multiple seeds

• multiple health-system configurations:

  • disable_and_reject_all=True
  • mode=0, mode=2

And posting errors to the comments.

[I think this might be done already?]

/calib

Generate the standard set of graphics that we are using to assess the calibration of the model to demographic and epidemiological data.

1) Run long_run_all_diseases.py

2) Run the scripts in this folder and post the figures from each to the comment thread:

(One of which analysis_hsi_descriptions.py under development in https://github.com/UCL/TLOmodel/pull/436).

(Not sure if/how these scripts need to be modified to make then amenable this kind of usage).

/hsi-effects The aim here is to produce a summary figure of the effect of disallowing each TREATMENT_ID from being run, compared to when all TREATEMENT_IDs can run, in terms of the different in deaths/dalys by age/sex. Code is under development: https://github.com/UCL/TLOmodel/issues/515

[tbhallett]

Noting here that we now have a lot of this functionality available on the separate dev server.