2_process.R

source('2_process/src/data_utils.R')

p2_targets <- list(
  # Confirming raw data matches `p1_unc_stats` from SB
  tar_target(p2_unc_agg_summary,
             p1_unc_agg |>
               group_by(dimension, determinant) |>
               summarize(across(c(contains('related'),
                                  contains('unknown'),
                                  contains('significant')),
                                list(total = ~sum(.x, na.rm=TRUE)))) |> 
               mutate(evidence_val = positively_related_total + negatively_related_total +
                        unrelated_total + unknown_direction_total)
             ),
  # Based on metadata:
  # Amt of evidence: Small = total_studies < 5; Medium = total_studies 5-9; Large,total_studies = > 9
  # Amt of agreement: Low = < 50% of models; Medium = >50% & <74% of models; High = >74% of models; NA if the level of agreement could not be calculated as indicator was measured only once.
  tar_target(p2_top_trend_stats,
             p2_unc_agg_summary |>
               dplyr::select(dimension, determinant, #indicator, 
                      positively_related_total, negatively_related_total, unrelated_total, 
                      unknown_direction_total) |>
               pivot_longer(!c(dimension,determinant)) |>
               group_by(dimension, determinant) |>
               # for each indicator find the maximum % of studies in agreement 
               # across the significance categories. 
               slice_max(value) |>
               rename(sig_name = name, sig_value = value)
             ),
# Join `p2_unc_agg_summary` to top trends to get percentages of agreement and evidence
  tar_target(p2_unc_agg_summary_csv,
             p2_unc_agg_summary |>
               left_join(p2_top_trend_stats) |>
               # level of agreement is the max percent of studies in agreement
               dplyr::mutate(level_agreement = 100*(sig_value/evidence_val)) |>
               readr::write_csv('2_process/out/indicator_uncertainty.csv')
             ),

  tar_target(p2_indicators,
             p1_unc_stats |>
               distinct(dimension, determinant, indicator)
             ),
  # Process census data for variables of interest
  # B01003_001 =  Total Population 
  # B19013_001 = Median Household Income in the Past 12 Months (in 2022 Inflation-Adjusted Dollars)
  # B02001_003 =  Estimate!!Total:!!Black or African American alone
  # B03001_003 = Estimate!!Total:!!Hispanic or Latino:
  # B01001_002 = Estimate!!Total:!!Male:
  # B01001_026 = Estimate!!Total:!!Female:
  tar_target(p2_census_layers,
             list("B01003_001", "B19013_001", "B02001_003",
                  "B03001_003", "B01001_002", "B01001_026")
             ),
  tar_target(p2_census_data,
             get_census_data(geography = 'county', variable = p2_census_layers,
                             states = p1_census_states, year = 2022, proj = p1_proj),
             pattern = map(p2_census_layers),
             iteration = "list"
             ),
  tar_target(p2_tot_pop,
             p2_census_data[[1]] |>
               st_drop_geometry() |>
               rename(tot_pop = estimate)),
  # Add % of total population col to each census layer
  tar_target(p2_perc_census_layers_sf,
             process_perc(tot_var = p2_census_data,
                          tot_pop = p2_tot_pop),
             pattern = map(p2_census_data),
             iteration = "list")
  )