updates from paper analyses for HIV, TB and malaria

[tdm32]

Updates include:

palliative care HSI for HIV
palliative care HSI for TB
priority ranking spreadsheet updated to include new EndOfLifeCare HSIs
updated TB diagnostic referral pathway, moving all xpert testing to facility level 2 (as consumables for xpert not available at level 1a and 1b)
remove unused parameters
update data in ResourceFiles
remove shorter paediatric treatment for TB as not commonly used in practice
add consistent repeat healthcare seeking behaviour cap for HIV, TB and malaria HSIs
remove parameters and code relating to scenario setting (not required for master, only for analyses)
include data_parameter for TB to match updated data (ending 2022)
new TbDecideDeath feature schedules hospital stay prior to death and determines probability of death via. linear model
HSI_Tb_ScreeningAndRefer accepts argument facility_level
HSI_ClinicaDiagnosis is now separate (was previously just a diagnostic test) meaning we can track the HR requirements
additional properties in HIV: hv_on_cotrimoxazole, hv_date_treated
schedule hospital stay prior to AIDS death
These changes were all implemented for branches used for the analysis in the recent two papers and so outputs have been checked and are all consistent with expectations.

[tdm32]

@tbhallett I have an error occurring in test_analysis: test_get_parameter_functions
Dataframe or series not of the expected size and shape:Tb:rate_testing_active_tb >> updated_value= year ... coverage_ipt_paediatric_contacts

I can see it's down to the data contained in ResourceFile_Improved_Healthsystem_And_Healthcare_Seeking but I don't want to edit this without Bingling's input. Let me know what you'd prefer me to do.

[tdm32]

@tbhallett Hi Tim, this error is now occurring but I'm not sure what the correct fix would be:

test_analysis.py:315: in test_get_parameter_functions
    assert name in sim.modules[module].parameters, f"Parameter not recognised: {module}:{name}."
E   AssertionError: Parameter not recognised: CareOfWomenDuringPregnancy:prob_intervention_delivered_ifa.
E   assert 'prob_intervention_delivered_ifa' in {'effect_of_ifa_for_resolving_anaemia': [0.7, 0.7], 'prob_adherent_ifa': [0.37, 0.34], 'prob_delivery_modes_ec': [[0, 0, 1], [0, 0, 1]], 'prob_delivery_modes_spe': [[0.05, 0.05, 0.9], [0.05, 0.05, 0.9]], ...}
E    +  where {'effect_of_ifa_for_resolving_anaemia': [0.7, 0.7], 'prob_adherent_ifa': [0.37, 0.34], 'prob_delivery_modes_ec': [[0, 0, 1], [0, 0, 1]], 'prob_delivery_modes_spe': [[0.05, 0.05, 0.9], [0.05, 0.05, 0.9]], ...} = <tlo.methods.care_of_women_during_pregnancy.CareOfWomenDuringPregnancy object at 0x7fcc589162c0>.parameters

I had updated the spreadsheet resources/ResourceFile_Improved_Healthsystem_And_Healthcare_Seeking.xlsx
as there is now updated data in ResourceFile_TB which was causing a conflict. No other data was changed.

Is it possible that the parameter CareOfWomenDuringPregnancy:prob_intervention_delivered_ifa was removed at some point but still exists in ResourceFile_Improved_Healthsystem_And_Healthcare_Seeking.xlsx? Should it still be there?

[tdm32]

@tbhallett one test is failing in test_healthsystem:test_all_treatment_ids_defined_in_priority_policies
where treatment ID PostnatalCare_Comprehensive is not defined in the priority policies but is in clean_set_of_filtered_treatment_ids. It occurs in HSI_Labour_ReceivesComprehensiveEmergencyObstetricCare. Should this be included in the priority policies?

[tbhallett]

@tbhallett This PR does contain the buggy commit which includes those changes around the malaria do_at_generic_first_appt function. When this is resolved, it may cause conflicts in this branch.

That's ok, we can still go ahead and merge this branch as the fix being developed targets master

[tdm32]

@tbhallett This is ready for review now