R/global.R

In shinymrp: Interface for Multilevel Regression and Poststratification

#' @title Global list of shiny module IDs
#' @noRd
#' @keywords internal
#' @keywords internal
module_ids <- list(
  home = "home",
  analyze = list(
    upload = "analyze_upload",
    visualize = "analyze_visualize",
    model = "analyze_model",
    result = "analyze_result"
  ),
  learn = list(
    interface = "learn_interface",
    preprocess = "learn_preprocess",
    mrp = "learn_mrp"
  ),
  about = "about",
  persist = "persist"
)

#' @title Global Constants
#' @description A global list containing paths, UI styles, and other constants used throughout the application.
#' @noRd
#' @keywords internal
#' @keywords internal
.const <- function() {
  list(
    plot = list(
      point_size = 3.5,
      errorbar_size = 0.8,
      errorbar_width = 0,
      raw_color = "darkblue",
      yrep_color = "darkorange",
      mrp_color = "darkorange",
      ui = list(
        plot_height = 550,
        subplot_height = 300,
        map_height = 700
      ),
      save = list (
        width = 18,
        height = 8,
        dpi = 300,
        units = "in"
      )
    ),
    ui = list(
      style = list(
        global = "
          .navbar-brand {
            font-size: 1.5rem;
          }
          
          .navbar-nav .nav-link {
            font-size: 1.04rem;
          }
          
          .nav-item .nav-link {
            font-size: 1.0rem;
          }"
      ),
      preview_size = 100,
      format = list(
        date = "%b%d\n%Y",
        data = c(".csv")
      ),
      model = list(
        max_models = 5,
        iter_range = c(100, 5000),
        chain_range = c(1, 8)
      ),
      loading_types = c("fit", "pstrat", "loo", "setup", "init", "wait"),
      guide_sections = c("workflow", "upload", "model_spec", "model_fit"),
      geo_view = c("map", "line_scatter"),
      animation = list(
        duration = 1000,
        delay = 100
      ),
      plot_selection = list(
        vis_main = list(
          binomial = c(
            "Individual Characteristics" = "indiv",
            "Geographic Characteristics" = "geo",
            "Positive Response Rate" = "outcome"
          ),
          normal = c(
            "Individual Characteristics" = "indiv",
            "Geographic Characteristics" = "geo",
            "Outcome Average" = "outcome"
          )
        ),
        indiv = c(
          "Sex" = "sex",
          "Race" = "race",
          "Age" = "age",
          "Education" = "edu"
        ),
        geo = c(
          "Sample Size" = "sample"
        ),
        geo_covar = c(
          "Education" = "college",
          "Poverty" = "poverty",
          "Employment" = "employment", 
          "Income" = "income",
          "Urbanicity" = "urbanicity",
          "ADI" = "adi"
        ),
        outcome = c(
          "Overall" = "overall",
          "By Geography" = "by_geo"
        ),
        summary = c(
          "Highest" = "max",
          "Lowest" = "min"
        ),
        subgroup = c(
          "Sex" = "sex",
          "Race" = "race",
          "Age" = "age",
          "Education" = "edu",
          "Geography" = "geo"
        )
      ),
      use_case_labels = list(
        covid = "Time-varying: COVID",
        poll = "Cross-sectional: Poll",
        timevar_general = "Time-varying: General",
        static_general = "Cross-sectional: General"
      )
    ),
    args = list(
      acs_years = 2018:2023,
      effect_types = c("fixed", "varying", "interaction"),
      family = c("binomial", "normal"),
      summary_types = c("max", "min"),
      time_freq = c("week", "month", "year")
    ),
    vars = list(
      pstrat = c("sex", "race", "age", "edu", "county", "state"),
      indiv  = c("sex", "race", "age", "edu", "time"),
      demo   = c("sex", "race", "age", "edu"),
      covar  = c("college", "poverty", "employment", "income", "urbanicity", "adi"),
      geo    = c("zip", "county", "state"),
      geo2   = c("county", "state"),
      time   = c("time", "date"),
      ignore = c("date", "total", "positive", "outcome")
    ),
    default_priors = list(
      intercept = "normal(0, 5)",
      fixed = "normal(0, 3)",
      varying = "normal(0, 3)",
      interaction = "normal(0, 3)",
      global_scale = "cauchy(0 , 1)",
      local_scale = "normal(0, 1)",
      icar_scale = "normal(0, 1)",
      bym2_scale = "normal(0, 3)",
      bym2_rho = "beta(0.5, 0.5)"
    ),
    custom_priors = c("structured", "icar", "bym2")
  )
}