Add Costing Resource Files and Scripts

[sakshimohan]

This PR adds the following -

1. Costing utility functions (cost_estimation.py) and related resource files (~/resources/costing/)
2. Analysis scripts for two manuscripts - 1. Method for costing a health system using a Health Systems Model - updated version submitted in May 2025; 2. System-Wide Investments Enhance HIV, TB and Malaria Control in Malawi and Deliver Greater Health Impact - updated version submitted in May 2025.

For more information, see ~src/scripts/costing/README.md.

Note that these files were extracted from a previous [PR #1279], which was closed due to conflicts with master.

Possible future improvements to the costing module are listed under Issue #1635

[sakshimohan]

Hi @tbhallett. I've now closed the old costing PR (PR #1279) and moved the relevant scripts and resource files here. This PR is now ready for review.

I have added a README.md file to describe the utility functions for costing. Happy to move this to a Wiki if that's preferable.

One test is failing. It seems to be related to the test files (not part of the costing scripts):

(TLOmodel) sm2511@UOY25M013 TLOmodel % ruff check tests --show-fixes
tests/bitset_handler/conftest.py:10:1: I001 [*] Import block is un-sorted or un-formatted
tests/bitset_handler/test_bitset_pandas_dtype.py:1:1: I001 [*] Import block is un-sorted or un-formatted
tests/bitset_handler/test_bitset_set_like_interactions.py:4:1: I001 [*] Import block is un-sorted or un-formatted
tests/test_alri.py:2:1: I001 [*] Import block is un-sorted or un-formatted
tests/test_analysis.py:1:1: I001 [*] Import block is un-sorted or un-formatted
tests/test_antenatal_interventions.py:1:1: I001 [*] Import block is un-sorted or un-formatted
tests/test_basic_sims.py:1:1: I001 [*] Import block is un-sorted or un-formatted
.....

[tbhallett]

Hi @tbhallett. I've now closed the old costing PR (PR #1279) and moved the relevant scripts and resource files here. This PR is now ready for review.

I have added a README.md file to describe the utility functions for costing. Happy to move this to a Wiki if that's preferable.

One test is failing. It seems to be related to the test files (not part of the costing scripts):

(TLOmodel) sm2511@UOY25M013 TLOmodel % ruff check tests --show-fixes
tests/bitset_handler/conftest.py:10:1: I001 [*] Import block is un-sorted or un-formatted
tests/bitset_handler/test_bitset_pandas_dtype.py:1:1: I001 [*] Import block is un-sorted or un-formatted
tests/bitset_handler/test_bitset_set_like_interactions.py:4:1: I001 [*] Import block is un-sorted or un-formatted
tests/test_alri.py:2:1: I001 [*] Import block is un-sorted or un-formatted
tests/test_analysis.py:1:1: I001 [*] Import block is un-sorted or un-formatted
tests/test_antenatal_interventions.py:1:1: I001 [*] Import block is un-sorted or un-formatted
tests/test_basic_sims.py:1:1: I001 [*] Import block is un-sorted or un-formatted
.....

Thanks @sakshimohan -- will review this very soon

[sakshimohan]

I've now updated ResourceFile_Consumables_Inflow_Outflow_Ratio.csv to match with that in PR #1367. (Commit ec89299)

[tbhallett]

This is the key file containing all the functions that we expect to be re-used and which we want to make sure always works.

Given this, I propose moving this file to src\tlo\analysis\

[tbhallett]

here, and elsewhere, arguments are named with the leading underscore.

Whilst this all works perfectly, the convention is for "_" to be used for private members, and so it's unusual to have them as argument names.

https://betterstack.com/community/questions/what-does-underscores-before-object-name-mean-in-python/

So, to improve future readability, I would propose a refactor to just rename the arguments to be without the leading underscore.

[tbhallett]

Noting here that this change does cause a fail on test_simulation (comparing logs between interrupted and non-interrupted simulations). Since this PR does not touch any methods files, I do not know what is causing this. It also seems to pass sometimes when running locally: in particular, when running that one test stand-alone. This makes me thing its do with the 'simulation 'factory'.

[tamuri]

Looks like it's test_continuous_and_interrupted_simulations_equal. Something has slipped through... 😰

[tamuri]

Some unresolved comments above - are you happy to merge and deal with those in a later PR?

[tbhallett]

Some unresolved comments above - are you happy to merge and deal with those in a later PR?

yes, we've agreed that those points can be recorded as an issue for future consideration.