Skip to contents

Get ITHIM-results into correct format for VoI analysis

Source: R/extract_data_for_voi.R
extract_data_for_voi.Rd

This function extracts the relevant information from the multi_city_ithim object and gets the results into the correct format for further analysis.

Usage

extract_data_for_voi(
  NSCEN,
  NSAMPLES,
  SCEN_SHORT_NAME,
  outcome_age_groups,
  cities,
  multi_city_ithim
)

Arguments

NSCEN

number of scenarios (not incl. baseline)

NSAMPLES

number of model runs per city

SCEN_SHORT_NAME

names of the scenarios (incl. baseline)

outcome_age_groups

outcome age groups as defined as input parameters to the model

cities

list of cities for which the model was run

multi_city_ithim

list containing the ithim model information including results for the various model runs

Value

ithim_results list with the following objects:

summary_ylls_df: dateframe with total ylls (median, 5th and 95th percentiles) per age group and city (plus combined results)

voi_data_all_df: dataframe for all cities with all outcomes for all model runs, age groups and disease and scenario combinations

yll_per_hundred_thousand: yll per 100,000 people for each city, outcome age category, model run and disease and scen combination

yll_per_hundred_thousand_stats: total ylls per 100,000 (median, 5th and 95th percentiles) as sum across all disease per outcome age group, scenario and city (plus combined results)

outcome: total yll outcome for all outcome age categories per city and scenario and disease combination, also combined city result (sum)

Details

The function performs the following steps:

  • by looping through the cities:

    • calculate average outcome (yll) per person in the population considered by the model

    • calculate the total ylls per 100 000 for each outcome age category, scenario and disease combination and model run

    • calculate total yll outcome across all outcome age categories per city and scenario and disease combinations

    • create one dataframe for all cities with all outcomes for all model runs, age groups and disease and scenario combinations

  • compute yll per hundred thousand by outcome age group by summing across all diseases (double counting!) by city and scenario and also summing across all cities

  • create one dateframe with total ylls (median, 5th and 95th percentiles) per age group and city (plus combined results as sum across all cities)