R/plot_q24_items_owned.R

In Rare-Technology/HHS_Dashboard: PKG_TITLE

prep_q24_items_owned <- function(.data){


  hhs_Q24 <- .data[, c(
    "24a_item_radio_no",
    "24b_item_tv_no",
    "24c_item_satellite_no",
    "24d_item_phone_no",
    "24e_item_washing_maching_no",
    "24f_item_generator_no",
    "24g_item_fridge_no",
    "24h_item_motorboat_no",
    "24i_item_outboard_no",
    "24j_item_inboard_no",
    "24k_item_sailboat_no",
    "24l_item_bicycle_no",
    "24m_item_motorcycle_no",
    "24n_item_car_no",
    "24o_item_internet_no",
    "24p_item_other_no"
  )]
  #replace NA with 0s
  hhs_Q24[is.na(hhs_Q24)] <- 0
  hhs_Q24_01_items <- hhs_Q24
  
  #Convert matrix to 0 and 1 (ignore wanrning)
  hhs_Q24_01_items[hhs_Q24_01_items > 0] <- 1
  
  hhs_Q24_01 <- data.frame(hhs_Q24_01_items, 
                           ma_name = .data$maa)
  
  #Estimate Proportion of community members with each asset
  prop <- function(x) {
    round(sum(x) / length(x) * 100, 1)
  }
  
  assets_prop <-
    aggregate(
      cbind(
        hhs_Q24_01[, 1],
        hhs_Q24_01[, 2],
        hhs_Q24_01[, 3],
        hhs_Q24_01[, 4],
        hhs_Q24_01[, 5],
        hhs_Q24_01[, 6],
        hhs_Q24_01[, 7],
        hhs_Q24_01[, 8],
        hhs_Q24_01[, 9],
        hhs_Q24_01[, 10],
        hhs_Q24_01[, 11],
        hhs_Q24_01[, 12],
        hhs_Q24_01[, 13],
        hhs_Q24_01[, 14],
        hhs_Q24_01[, 15],
        hhs_Q24_01[, 16]
      ) ~ ma_name,
      data = hhs_Q24_01,
      FUN = prop
    )
  
  N_assets <- tapply(hhs_Q24_01[, 1], hhs_Q24_01[, 17], length)
  
  #Rename columns
  colnames(assets_prop) <-
    c(
      "MA name",
      "Radio",
      "TV",
      "Satellite",
      "Phone",
      "Washing_Machine",
      "Generator",
      "Fridge",
      "Motorboat",
      "Outboard",
      "Inboard",
      "Sailboat",
      "Bicycle",
      "Motorcycle",
      "Car",
      "Internet",
      "Other"
    )
  
  df <- tidyr::pivot_longer(assets_prop, cols=colnames(assets_prop)[-1])
  df <- df %>% dplyr::rename(Item=name, Proportion=value)
  df$Item <- factor(df$Item,
                    levels = c('Internet', 'Phone', 'Radio', 'Satellite', 'TV',
                               'Fridge', 'Generator', 'Washing_Machine',
                               'Inboard', 'Motorboat', 'Outboard', 'Sailboat',
                               'Bicycle', 'Car', 'Motorcycle',
                               'Other'))
  df$color <- map_color(df$Proportion)
  
  df
  # ## Assest by MA
  # assets_prop_totals <-
  #   rbind(
  #     assets_prop,
  #     c(
  #       NA,
  #       compute_summary_line(assets_prop[[2]], 1),
  #       compute_summary_line(assets_prop[[3]], 1),
  #       compute_summary_line(assets_prop[[4]], 1),
  #       compute_summary_line(assets_prop[[5]], 1),
  #       compute_summary_line(assets_prop[[6]], 1),
  #       compute_summary_line(assets_prop[[7]], 1),
  #       compute_summary_line(assets_prop[[8]], 1),
  #       compute_summary_line(assets_prop[[9]], 1),
  #       compute_summary_line(assets_prop[[10]], 1),
  #       compute_summary_line(assets_prop[[11]], 1),
  #       compute_summary_line(assets_prop[[12]], 1),
  #       compute_summary_line(assets_prop[[13]], 1),
  #       compute_summary_line(assets_prop[[14]], 1),
  #       compute_summary_line(assets_prop[[15]], 1),
  #       compute_summary_line(assets_prop[[16]], 1),
  #       compute_summary_line(assets_prop[[17]], 1)
  #     )
  #   )
  # ## Total number of assets
  # hhs_Q24$assets_no <- rowSums(hhs_Q24_01[, 1:16])
  # hhs_Q24$maa <- .data$maa
  # 
  # ## mean number of assets per hhs in each MA
  # Q24_summary_bind <-
  #   data.frame(cbind(N = N_assets, Assets_number = round(
  #     tapply(hhs_Q24$assets_no, hhs_Q24$maa, mean), 2
  #   )))
  # 
  # Q24_summary <-
  #   rbind(Q24_summary_bind, "Mean ± SE" = c(
  #     sum(Q24_summary_bind$N),
  #     compute_summary_line(Q24_summary_bind$Assets_number, 2)
  #   ))
  # Q24_summary <- tibble::rownames_to_column(Q24_summary, "MA name")
  # colnames(Q24_summary) <- c("MA name", "N", "Proportion (%)")
  # Q24 <- clean_plot_data(Q24_summary)
  # colnames(Q24) <- c("MA name", "N", "Average")
  # 
  # Q24
}

map_color <- Vectorize(function(x) {
  if (x >= 50) {
    'black'
  } else {
    'white'
  }
})

plot_q24_items_owned <- function(.data, ...){


  .data_plot <- prep_q24_items_owned(.data)
  
  
  
  title <- 'Assets owned by households'
  title <- stringr::str_wrap(title, width = 65)
  subtitle <- 'Proportion of households (%)'
  
  .data_plot <- .data_plot %>% 
  dplyr::mutate(
    `MA name` := factor(`MA name`)
  )
  
  .data_plot <- .data_plot %>% 
  dplyr::mutate(
    `MA name` := forcats::fct_rev(`MA name`)
  )
  
  
  p <- ggplot(data = .data_plot,
       aes(x = Item, y = `MA name`, fill = Proportion)) +
  geom_tile() + 
  geom_text(aes(label = Proportion, color=Proportion >= 50),
            show.legend = c(TRUE, FALSE)) +
  scale_fill_viridis_c(option='magma') +
  scale_color_manual(values=c('white', 'black')) +
  labs(title = title, subtitle=subtitle) +
  theme_rare() +
  theme(
    axis.text.x = element_text(angle=30, hjust=1),
    axis.ticks.x = element_line(),
    panel.grid.major.x = element_blank(),
    axis.title.y = element_blank()
  )
  
  result <- list(
    plot = p,
    data = .data_plot
  )
}