R/viz_plots.R

In randyyee/SaviR: ITF Analytics

#~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

#' @title plot_epicurve
#' @description (EPI WEEK) Visualize epi curve by epi-weeks (Monday-Sunday) and by WHO region(s), State region(s), or Income levels.
#'
#' @param df A dataframe with the following: date, new_cases and one of these columns for by_cat: who_region, state_region, or incomelevel_value.
#' Produces an epi curve, stacked bar plot for each epi-week (Monday-Sunday).
#' @param by_cat = "WHO Region" (default), "State Region" or "Income Level"
#' @param legend Default "in" - position legend inside the plot area.
#' @param transparent Default TRUE - returns a transparent plot.
#'
#' @importFrom magrittr `%>%`
#'
#' @export

plot_epicurve <- function(df, by_cat = "WHO Region", legend = "in", transparent = T) {

  if(grepl("WHO", by_cat, fixed = TRUE)) {
    cat_values <- c("AMRO", "EURO", "SEARO", "EMRO", "AFRO", "WPRO")
    cat_names  <- c("Americas", "Europe", "Southeast Asia", "Eastern Mediterranean", "Africa", "Western Pacific")
    cat_colors <- c("#aa001e", "#e7b351", "#00818a", "#d26230", "#005e70", "#d4ece8")
    df_c       <- df %>% dplyr::mutate(cat = factor(who_region,levels = cat_values))
  } else if(grepl("State", by_cat, fixed = TRUE)) {
    cat_values <- c("East Asia and the Pacific",
                    "Europe and Eurasia",
                    "Near East (Middle East and Northern Africa)",
                    "South and Central Asia",
                    "Sub-Saharan Africa",
                    "Western Hemisphere",
                    "US",
                    "None-state")
    cat_names  <- c("East Asia and the Pacific",
                    "Europe and Eurasia",
                    "Near East (Middle East and North Africa)",
                    "South and Central Asia",
                    "Sub-Saharan Africa",
                    "Western Hemisphere (not incl US)",
                    "US",
                    "None-state")
    cat_colors <- c("#d00000", "#ffba08", "#3f88c5", "#032b43", "#136f63", "#a5c651", "#d64550", "#808080")
    df_c       <- df %>% dplyr::mutate(cat = factor(state_region,levels = cat_values))
  } else if(grepl("Income", by_cat, fixed = TRUE)) {
    cat_values <- c("High income", "Upper middle income", "Lower middle income", "Low income", "Not classified")
    cat_names  <- cat_values
    cat_colors <- c("#045a8d", "#74a9cf", "#fdbb84", "#d7301f", "#808080")
    df_c       <- df %>% dplyr::mutate(cat = factor(incomelevel_value,levels = cat_values))
  }
  col_master <- data.frame(cat_values, cat_names, cat_colors)

  if(length(unique(df_c$cat)) > 1) {
    category_color_labels <- col_master$cat_names
    category_color_values <- col_master$cat_colors
    gtitle                <- paste0("Confirmed COVID-19 Cases by Week of Report and ", by_cat)
  } else {
    category_color_labels <- col_master[cat_values == as.character(unique(df_c$cat)), ]$cat_names
    category_color_values <- col_master[cat_values == as.character(unique(df_c$cat)), ]$cat_colors
    gtitle                <- paste0("Confirmed COVID-19 Cases - ", category_color_labels)
  }

  g <- ggplot2::ggplot(data    = df_c,
                       mapping = aes(x    = lubridate::floor_date(date, "week", week_start = 1),
                                     y    = new_cases,
                                     fill = cat)) +
    ggplot2::geom_bar(position = "stack",
                      stat     = "identity",
                      alpha    = 0.9) +
    ggplot2::labs(title    = gtitle,
                  subtitle = paste0(str_squish(format(min(df_c$date, na.rm = T), "%B %e, %Y")), " - ",
                                    str_squish(format(max(df_c$date, na.rm = T), "%B %e, %Y"))),
                  fill     = by_cat) +
    ggplot2::ylab("Weekly Cases") +
    ggplot2::xlab("Week of Reporting") +
    ggplot2::scale_x_date(limits = c(lubridate::floor_date(min(df_c$date, na.rm = T)-7, "week", week_start = 1),
                                     lubridate::floor_date(max(df_c$date, na.rm = T)+7, "week", week_start = 1)),
                          breaks = seq.Date(from = as.Date(lubridate::floor_date(min(df_c$date, na.rm = T), "week", week_start = 1)),
                                            to   = as.Date(lubridate::floor_date(max(df_c$date, na.rm = T)+7, "week", week_start = 1)),
                                            by   = "4 weeks"),
                          date_labels = "%e\n%b",
                          expand      = c(0, 0)) +
    ggplot2::scale_y_continuous(expand = c(0, 5000),
                                labels = scales::comma) +
    ggplot2::scale_fill_manual(values = category_color_values,
                               labels = category_color_labels) +
    ggplot2::theme_classic() +
    ggplot2::theme(plot.title      = ggplot2::element_text(size  = 17, face = "bold", family = "Calibri"),
                   axis.text       = ggplot2::element_text(size  = 10, family = "Calibri"),
                   axis.title      = ggplot2::element_text(size  = 12, family = "Calibri"),
                   legend.title    = ggplot2::element_text(size  = 12, face = "bold", family = "Calibri"),
                   legend.text     = ggplot2::element_text(size  = 9,  family = "Calibri"),
                   legend.key.size = unit(0.5, 'cm')) +
    ggplot2::guides(fill = ggplot2::guide_legend(overide.aex  = list(size = 9)))

  if(length(unique(df_c$cat)) > 1) {
    if(legend == "in") {
      g <- g + ggplot2::theme(legend.justification = c("left","top"),
                              legend.position = c(0,1))
    } else {
      g <- g + ggplot2::theme(legend.position = legend)
    }
  } else {
    g <- g + ggplot2::theme(legend.position = "none")
  }

  if(transparent == T){
    return(g +
             ggplot2::theme(panel.background  = ggplot2::element_rect(fill = "transparent"),
                            plot.background   = ggplot2::element_rect(fill = "transparent"),
                            panel.grid        = ggplot2::element_blank(),
                            legend.background = ggplot2::element_rect(fill = "transparent")))
  }else{
    return(g)
  }

}

#~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

#' @title plot_epicurve_ind
#' @description (DAILY) Visualize epi curve by cases and deaths.
#' Default viz for individual countries.
#'
#' @param df A dataframe with the following: country, date, cases and/or deaths
#' @param type Default cases.
#' @param incidence Default TRUE. Specify inputs are incidence values or not.
#'
#' @importFrom magrittr `%>%`
#'
#' @export

plot_epicurve_ind <- function(df, type = "cases", incidence = T){

  if(!type %in% c("cases", "deaths")){
    stop("Wrong Type! You must use either cases or deaths!")
  }

  if(!incidence %in% c(T, F)){
    stop("Wrong Incidence! You must use either TRUE or FALSE!")
  }

  if(incidence == F){
    df %>%
      ggplot2::ggplot(aes(x = date, y = if(type == "cases") {cases} else {deaths})) +
      ggplot2::geom_bar(stat = "identity", alpha = 0.9, fill = if(type == "cases") {"dodgerblue4"} else {"red4"}) +
      ggplot2::theme_classic() +
      ggplot2::ylab(if(type == "cases") {"Daily Cases"} else {"Daily Deaths"}) +
      ggplot2::xlab("Date of Reporting") +
      ggplot2::scale_x_date(limits = c(lubridate::floor_date(min(df$date, na.rm = T)-7, "week", week_start = 1),
                                       lubridate::floor_date(max(df$date, na.rm = T)+7, "week", week_start = 1)),
                            breaks = seq.Date(from = as.Date(lubridate::floor_date(min(df$date, na.rm = T), "week", week_start = 1)),
                                              to   = as.Date(lubridate::floor_date(max(df$date, na.rm = T)+7, "week", week_start = 1)),
                                              by   = "3 weeks"),
                            date_labels  = "%d\n%b") +
      ggplot2::scale_y_continuous(labels = comma,
                                  expand = expansion(mult = c(0.01, .1))) +
      ggplot2::labs(title    = if(type == "cases") {paste0("COVID-19 Cases: ", unique(df$country))} else {paste0("COVID-19 Deaths:", unique(df$country))},
                    subtitle = paste0(format(min(df$date, na.rm = T), "%B %d, %Y"), " - ", format(max(df$date, na.rm = T), "%B %d, %Y"))) +
      ggplot2::theme(plot.title   = ggplot2::element_text(size = 14, face = "bold", family = "Calibri"),
                     axis.text    = ggplot2::element_text(size = 10, family = "Calibri"),
                     axis.title   = ggplot2::element_text(size = 12, family = "Calibri"),
                     legend.title = ggplot2::element_text(size = 12, face = "bold", family = "Calibri"),
                     legend.text  = ggplot2::element_text(size = 9, family = "Calibri"))
  } else {
    df %>%
      ggplot2::ggplot(aes(x = date, y = if(type == "cases") {cases} else {deaths})) +
      ggplot2::geom_bar(stat = "identity", alpha = 0.9, fill = if(type == "cases") {"dodgerblue4"} else {"red4"}) +
      ggplot2::theme_classic() +
      ggplot2::ylab(if(type == "cases") {"Daily Cases per 100,000 People"} else {"Daily Deaths per 100,000 People"}) +
      ggplot2::xlab("Date of Reporting") +
      ggplot2::scale_x_date(limits = c(lubridate::floor_date(min(df$date, na.rm = T)-7, "week", week_start = 1),
                                       lubridate::floor_date(max(df$date, na.rm = T)+7, "week", week_start = 1)),
                            breaks = seq.Date(from = as.Date(lubridate::floor_date(min(df$date, na.rm = T), "week", week_start = 1)),
                                              to   = as.Date(lubridate::floor_date(max(df$date, na.rm = T)+7, "week", week_start = 1)),
                                              by   = "3 weeks"),
                            date_labels  = "%d\n%b") +
      ggplot2::scale_y_continuous(labels = comma,
                                  expand = expansion(mult = c(0.01, .1))) +
      ggplot2::labs(title    = if(type == "cases") {paste0("COVID-19 Cases per 100,000 People: ", unique(df$country))} else {paste0("COVID-19 Deaths per 100,000 People: ", unique(df$country))},
                    subtitle = paste0(format(min(df$date, na.rm = T), "%B %d, %Y"), " - ", format(max(df$date, na.rm = T), "%B %d, %Y"))) +
      ggplot2::theme(plot.title   = ggplot2::element_text(size = 14, face = "bold", family = "Calibri"),
                     axis.text    = ggplot2::element_text(size = 8,  family = "Calibri"),
                     axis.title   = ggplot2::element_text(size = 10, family = "Calibri"),
                     legend.title = ggplot2::element_text(size = 12, face = "bold", family = "Calibri"),
                     legend.text  = ggplot2::element_text(size = 9,  family = "Calibri"))
  }
}


#~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

#' @title plot_epicurve_epidouble
#' @description (EPI WEEK) Visualize epi curve by cases and deaths.
#' Default viz for individual countries.
#'
#' @param df A dataframe with the following: country, weekdate, cases and deaths
#'
#' @importFrom magrittr `%>%`
#'
#' @export

plot_epicurve_epidouble <- function(df){

  ylim.prim <- c(min(df$case, na.rm = T),
                 max(df$case, na.rm = T))

  ylim.sec  <- c(min(df$death, na.rm = T),
                 max(df$death, na.rm = T))

  b <- diff(ylim.prim)/diff(ylim.sec)
  a <- ylim.prim[1] - b * ylim.sec[1]

  ggplot2::ggplot(df) +
    ggplot2::geom_bar(aes(x = weekdate, y = case, color = "Cases"), stat = "identity", alpha = 0.9, fill = "lightblue") +
    ggplot2::geom_line(aes(x = weekdate, y = a + death * b, group = 1, color = "Deaths"), size = 1) +
    ggplot2::scale_color_manual(breaks = c("Cases", "Deaths"),
                                values = c("lightblue", "red")) +
    ggplot2::theme_classic() +
    ggplot2::scale_x_date(breaks       = c(by = "4 weeks"),
                          date_labels  = "%d\n%b") +
    ggplot2::scale_y_continuous("Weekly Cases", labels = comma,
                                sec.axis = sec_axis(~ (. - a)/b, name = "Weekly Deaths", labels = comma)) +
    ggplot2::xlab("Date of Reporting") +
    ggplot2::labs(title    = paste0("COVID-19: ", unique(df$country)),
                  subtitle = paste0("Week of:", format(min(df$weekdate, na.rm = T), "%B %d, %Y"), " - ", format(max(df$weekdate, na.rm = T), "%B %d, %Y"))) +
    ggplot2::theme(plot.title      = ggplot2::element_text(size = 16, face = "bold", family = "Calibri"),
                   axis.text       = ggplot2::element_text(size = 14, family = "Calibri"),
                   axis.title      = ggplot2::element_text(size = 14, family = "Calibri"),
                   legend.position = "top",
                   legend.key      = element_blank(),
                   legend.title    = ggplot2::element_blank(),
                   legend.text     = ggplot2::element_text(size = 12, family = "Calibri"))  +
    ggplot2::guides(color          = ggplot2::guide_legend(override.aes = list(fill = c("lightblue", NA))))
}


#~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

#' @title plot_epicurve_dailydouble
#' @description (DAILY) Visualize epi curve by cases and deaths.
#' Default viz for individual countries.
#'
#' @param df A dataframe with the following: country, date, cases and deaths
#'
#' @importFrom magrittr `%>%`
#'
#' @export

plot_epicurve_dailydouble <- function(df){

  ylim.prim <- c(min(df$case, na.rm = T),
                 max(df$case, na.rm = T))

  ylim.sec  <- c(min(df$death, na.rm = T),
                 max(df$death, na.rm = T))

  b <- diff(ylim.prim)/diff(ylim.sec)
  a <- ylim.prim[1] - b * ylim.sec[1]

  ggplot2::ggplot(df) +
    ggplot2::geom_bar(aes(x = date, y = case, color = "Cases"), stat = "identity", alpha = 0.9, fill = "lightblue") +
    ggplot2::geom_line(aes(x = date, y = a + death * b, group = 1, color = "Deaths"), size = 1) +
    ggplot2::scale_color_manual(breaks = c("Cases", "Deaths"),
                                values = c("lightblue", "red")) +
    ggplot2::theme_classic() +
    ggplot2::scale_y_continuous("Cases", labels = comma,
                                sec.axis = sec_axis(~ (. - a)/b, name = "Deaths", labels = comma)) +
    ggplot2::xlab("Date of Reporting") +
    ggplot2::labs(title    = paste0("COVID-19: ", unique(df$country)),
                  subtitle = paste0("Week of:", format(min(df$weekdate, na.rm = T), "%B %d, %Y"), " - ", format(max(df$weekdate, na.rm = T), "%B %d, %Y"))) +
    ggplot2::theme(plot.title      = ggplot2::element_text(size = 16, face = "bold", family = "Calibri"),
                   axis.text       = ggplot2::element_text(size = 14, family = "Calibri"),
                   axis.title      = ggplot2::element_text(size = 14, family = "Calibri"),
                   legend.position = "top",
                   legend.key      = element_blank(),
                   legend.title    = ggplot2::element_blank(),
                   legend.text     = ggplot2::element_text(size = 12, family = "Calibri"))  +
    ggplot2::guides(color          = ggplot2::guide_legend(override.aes = list(fill = c("lightblue", NA))))
}


#~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

#' @title plot_riskmatrix
#' @description Plot risk matrix.
#' @param df A dataframe with riskmatrix stats.
#'
#' @importFrom magrittr `%>%`
#'
#' @export

plot_riskmatrix <- function(df){
  ggplot2::ggplot(data = df, aes(x = percent_change_case, y = week_case_incidence)) +
    ggplot2::geom_point(aes(size = week_case, color = who_region), alpha=0.7) +
    ggplot2::scale_color_manual(values = c("#aa001e", "#e7b351", "#00818a", "#d26230", "#005e70", "#d4ece8"),
                                labels = c("Americas", "Europe", "Southeast Asia", "Eastern \nMediterranean", "Africa", "Western Pacific")) +
    ggrepel::geom_text_repel(aes(label = labels),
                             color              = 'black',
                             size               = 2.7,
                             min.segment.length = 0,
                             seed               = 42,
                             box.padding        = 0.6) +
    ggplot2::scale_size(name   = "Weekly Cases",
                        range  = c(2, 12),
                        breaks = c(100, 1000, 10000, 100000, 250000, 500000, 750000),
                        labels = scales::comma) +
    ggplot2::guides(color = guide_legend(override.aes = list(size = 6), order = 1)) +
    ggplot2::xlim(min(df$percent_change_case, na.rm = T), max(df$percent_change_case, na.rm = T)) +
    ggplot2::ylim(0,max(df$week_case_incidence, na.rm = T)) +
    ggplot2::xlab("% Change in Weekly Cases") + labs(color="WHO Region")+
    ggplot2::ylab("Average Daily Incidence per 100,000") +
    ggplot2::geom_vline(xintercept = 0,    color = 'gray50',    lty = 2) +
    ggplot2::geom_hline(yintercept = 0,    color = "green3",    linetype = "dashed") +
    ggplot2::geom_hline(yintercept = 1.0,  color = "goldenrod1",linetype = "dashed") +
    ggplot2::geom_hline(yintercept = 10.0, color = "orange2",   linetype = "dashed") +
    ggplot2::geom_hline(yintercept = 25.0, color = "red3",      linetype = "dashed") +
    ggplot2::annotate(geom = "text", x = -133, y = 0.6,  label = "< 1.0 per 100k",       color = "green3",    size = 3)+
    ggplot2::annotate(geom = "text", x = -125, y = 1.7,  label = "1.0 - 9.9 per 100k",   color = "goldenrod1",size = 3)+
    ggplot2::annotate(geom = "text", x = -122, y = 10.7, label = "10.0 - 24.9 per 100k", color = "orange2",   size = 3)+
    ggplot2::annotate(geom = "text", x = -133, y = 25.7, label = "25.0+ per 100k",       color = "red3",      size = 3)+
    ggplot2::theme_classic() +
    ggplot2::theme(axis.text     = element_text(size = 8, family = "Calibri"),
                   axis.title    = element_text(size = 10, family = "Calibri"),
                   legend.text   = element_text(size = 7, family = "Calibri"),
                   legend.title  = element_text(size = 9, family = "Calibri"),
                   plot.title    = element_text(size = 16, face = "bold", family = "Calibri"),
                   plot.subtitle = element_text(size = 11, family = "Calibri"),
                   plot.margin   = unit(c(0.5,0.5,0.5,0.5),"cm"),
                   plot.caption  = element_text(hjust = 0, size = 11, family = "Calibri")) +
    ggplot2::labs(title    = "Burden and Recent Trends",
                  subtitle = paste0("Average daily incidence per 100,000 population and 7-day percent change, by new cases in past 7 days\n",
                                    format(max(df$date)-13, "%B %d, %Y"), ' - ', format(max(df$date)-7, "%B %d, %Y"), ' to ',
                                    format(max(df$date)-6, "%B %d, %Y"), ' - ', format(max(df$date), "%B %d, %Y")),
                  caption  = "Notes:
      - Includes countries with a population greater than 10 million people and more than 100 cases in the last week
      - Countries with a population over 10 million are labeled if they are among the top ten highest countries for cases,
        incidence, or weekly percent change in cases.")
}


#~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

#' @title plot_vaxcoverage
#' @description Plot vaccination coverage (partial or fully) by WHO region(s), State region(s), or Income levels.
#' @param df A dataframe with vaccination stats.
#'
#' @importFrom magrittr `%>%`
#'
#' @export

plot_vaxcoverage <- function(df, type = "partial", by_cat = "State Region") {

  if(by_cat == "WHO Region") {
    cat_values <- c("AMRO", "EURO", "SEARO", "EMRO", "AFRO", "WPRO")
    cat_names  <- c("Americas", "Europe", "Southeast Asia", "Eastern Mediterranean", "Africa", "Western Pacific")
    cat_colors <- c("#aa001e", "#e7b351", "#00818a", "#d26230", "#005e70", "#d4ece8")
    df_c       <- df %>% dplyr::mutate(cat = factor(who_region,levels = cat_values))
  } else if(by_cat == "State Region") {
    cat_values <- c("East Asia and the Pacific",
                    "Europe and Eurasia",
                    "Near East (Middle East and Northern Africa)",
                    "South and Central Asia",
                    "Sub-Saharan Africa",
                    "Western Hemisphere",
                    "US",
                    "None-state")
    cat_names  <- c("East Asia and the Pacific",
                    "Europe and Eurasia",
                    "Near East (Middle East and North Africa)",
                    "South and Central Asia",
                    "Sub-Saharan Africa",
                    "Western Hemisphere (not incl US)",
                    "US",
                    "None-state")
    cat_colors <- c("#d00000", "#ffba08", "#3f88c5", "#032b43", "#136f63", "#a5c651", "#d64550", "#808080")
    df_c       <- df %>% dplyr::mutate(cat = factor(state_region,levels = cat_values))
  } else if(by_cat == "Income Level") {
    cat_values <- c("High income", "Upper middle income", "Lower middle income", "Low income", "Not classified")
    cat_names  <- cat_values
    cat_colors <- c("#045a8d", "#74a9cf", "#fdbb84", "#d7301f", "#808080")
    df_c       <- df %>% dplyr::mutate(cat = factor(incomelevel_value,levels = cat_values))
  }
  col_master <- data.frame(cat_values, cat_names, cat_colors)

  category_color_labels <- col_master$cat_names
  category_color_values <- col_master$cat_colors

  if(type == "partial") {
    df_c <- df_c %>%
      group_by(cat) %>%
      mutate(rank_people    = dense_rank(-people_vaccinated_per_hundred),
             rank_total     = dense_rank(-total_vaccinations)) %>%
      mutate(country_labels = case_when(rank_people %in% 1:3 ~ country,
                                        rank_total  %in% 1:3 ~ country)) %>%
      ungroup()
    ptitle <- paste0("People Vaccinated per 100 people by ", by_cat, ", ", format(max(df$date), "%B %d, %Y"))
    xlabel <- "People Vaccinated per 100"
    cap    <- "Notes:
    - People Vaccinated per 100: number of people who received at least one vaccine dose; does not represent
      percent of population fully vaccinated
    - Total vaccine doses administered: total doses given, does not represent number of people vaccinated
    - Countries are labeled such that within each group, labeled countries are those that are the top 3 ranking countries
      for people vaccinated per 100 and the top 3 ranking countries for total vaccine doses administered
    - Vaccine data are incomplete and data may be out of date"
  } else {
    df_c <- df_c %>%
      group_by(cat) %>%
      mutate(rank_people    = dense_rank(-people_fully_vaccinated_per_hundred),
             rank_total     = dense_rank(-total_vaccinations)) %>%
      mutate(country_labels = case_when(rank_people %in% 1:3 ~ country,
                                        rank_total  %in% 1:3 ~ country)) %>%
      ungroup()
    ptitle <- paste0("People Fully Vaccinated per 100 people by ", by_cat, ", ", format(max(df$date), "%B %d, %Y"))
    xlabel <- "People Fully Vaccinated per 100"
    cap    <- "Notes:
    - Total vaccine doses administered: total doses given, does not represent number of people fully vaccinated
    - Countries are labeled such that within each group, labeled countries are those that are the top 3 ranking countries
      for people fully vaccinated per 100 and the top 3 ranking countries for total vaccine doses administered
    - Vaccine data are incomplete and data may be out of date"
  }

  my_pal_vax <- function(range = c(3, 25)) {
    force(range)
    function(x) scales::rescale(x, to = range, from = c(0, 1))
  }

  ggplot2::ggplot(df_c, aes(x = if(type == "partial") {people_vaccinated_per_hundred}
                                else {people_fully_vaccinated_per_hundred},
                            y = cat)) +
    ggplot2::geom_point(aes(size = total_vaccinations, fill = cat),
                        shape = 21,
                        color = 'gray60',
                        alpha = 0.8) +
    ggrepel::geom_text_repel(aes(label = country_labels, point.size = total_vaccinations),
                             color              = "gray25",
                             min.segment.length = 0,
                             max.overlaps       = Inf,
                             size               = 3,
                             force              = 0.7,
                             force_pull         = 0.7,
                             direction          = "both",
                             box.padding        = 0.4,
                             point.padding      = 0)+
    ggplot2::continuous_scale(aesthetics = c("size", "point.size"), scale_name = "size", palette = my_pal_vax(),
                              labels = scales::comma, breaks = c(1000000, 50000000, 300000000, 750000000),
                              guide  = guide_legend(override.aes = list(label = "")),
                              name   = "Total vaccine \ndoses administered") +
    ggplot2::scale_fill_manual(name   = by_cat,
                               values = category_color_values,
                               labels = category_color_labels) +
    ggplot2::scale_x_continuous(name = xlabel) +
    ggplot2::scale_y_discrete(name = NULL) +
    ggplot2::guides(fill = guide_legend(reverse = TRUE, override.aes = list(size = 8))) +
    ggplot2::theme_bw() +
    ggplot2::labs(title = ptitle,
                  caption = cap,
                  legend.title  = element_text(size = 10, face = "bold", family = "Calibri")) +
    ggplot2::theme(plot.title   = element_text(size = 14, face = "bold", family = "Calibri"),
                   axis.title   = element_text(size = 12, family = "Calibri"),
                   plot.caption = element_text(hjust = 0, size = 12, family = "Calibri"))
}


#~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

#' @title plot_vaxcurve
#' @description Visualize vaccine coverage by date of reporting and by WHO region(s), State region(s), or Income levels.
#'
#' @param df A dataframe with the following: date, people_vaccinated_per_hundred or people_fully_vaccinated_per_hundred, and one of these columns for by_cat: who_region, state_region, or incomelevel_value.
#' @param type = "partial" (default) for partial vaccinated or "full" for fully vaccinated 
#' @param by_cat = "State Region" (default), "WHO Region" or "Income Level"
#' @param countries = "All" (default) for all countries or "AMC/AU" for AMC/AU countries (n=100)
#'
#' @importFrom magrittr `%>%`
#'
#' @export

plot_vaxcurve <- function(df, type = "partial", by_cat = "Dept. of State Region", countries = "All") {
  
  if(grepl("WHO", by_cat, fixed = TRUE)) {
    cat_values <- c("AMRO", "EURO", "SEARO", "EMRO", "AFRO", "WPRO")
    cat_names  <- c("Americas", "Europe", "Southeast Asia", "Eastern Mediterranean", "Africa", "Western Pacific")
    cat_colors <- c("#aa001e", "#e7b351", "#00818a", "#d26230", "#005e70", "#d4ece8")
    cat_lines  <- c("solid", "solid", "solid", "solid", "solid", "solid")
    df_c       <- df %>% dplyr::mutate(cat = factor(who_region,levels = cat_values))
  } else if(grepl("State", by_cat, fixed = TRUE)) {
    if(countries == "AMC/AU") {
      cat_values <- c("East Asia and the Pacific",
                      "Europe and Eurasia",
                      "Near East (Middle East and Northern Africa)",
                      "South and Central Asia",
                      "Sub-Saharan Africa",
                      "Western Hemisphere",
                      "None-state")
      cat_names  <- c("East Asia and the Pacific",
                      "Europe and Eurasia",
                      "Near East (Middle East and North Africa)",
                      "South and Central Asia",
                      "Sub-Saharan Africa",
                      "Western Hemisphere",
                      "None-state")
      cat_colors <- c("#d00000", "#ffba08", "#3f88c5", "#032b43", "#136f63", "#a5c651", "#808080")
      cat_lines  <- c("solid", "solid", "solid", "solid", "solid", "solid", "solid")
    } else {
      cat_values <- c("East Asia and the Pacific",
                      "Europe and Eurasia",
                      "Near East (Middle East and Northern Africa)",
                      "South and Central Asia",
                      "Sub-Saharan Africa",
                      "Western Hemisphere",
                      "US",
                      "None-state")
      cat_names  <- c("East Asia and the Pacific",
                      "Europe and Eurasia",
                      "Near East (Middle East and North Africa)",
                      "South and Central Asia",
                      "Sub-Saharan Africa",
                      "Western Hemisphere (not incl US)",
                      "US",
                      "None-state")
      cat_colors <- c("#d00000", "#ffba08", "#3f88c5", "#032b43", "#136f63", "#a5c651", "#aaaaaa", "#808080")
      cat_lines  <- c("solid", "solid", "solid", "solid", "solid", "solid", "dashed", "solid")
    }
    df_c       <- df %>% dplyr::mutate(cat = factor(state_region,levels = cat_values))
  } else if(grepl("Income", by_cat, fixed = TRUE)) {
    cat_values <- c("High income", "Upper middle income", "Lower middle income", "Low income", "Not classified")
    cat_names  <- cat_values
    cat_colors <- c("#045a8d", "#74a9cf", "#fdbb84", "#d7301f", "#aaaaaa")
    cat_lines  <- c("solid", "solid", "solid", "solid", "solid")
    df_c       <- df %>% dplyr::mutate(cat = factor(incomelevel_value,levels = cat_values))
  }
  col_master <- data.frame(cat_values, cat_names, cat_colors, cat_lines)
  
  category_labels       <- col_master$cat_names
  category_color_values <- col_master$cat_colors
  category_line_values  <- col_master$cat_lines
  if(type == "full") {
    gtitle <- "People fully vaccinated per 100 people"
  } else {
    gtitle <- "People with at least one vaccine dose per 100 people"
  }
  if(countries != "All") {
    gtitle <- paste0(gtitle, " in ", countries, " countries")
  }
  
  g <- ggplot2::ggplot(data    = df_c,
                       mapping = aes(x        = date,
                                     y        = if(type == "full") {people_fully_vaccinated_per_hundred_r} 
                                     else {people_vaccinated_per_hundred_r},
                                     colour   = cat,
                                     linetype = cat)) +
    ggplot2::geom_line() +
    ggplot2::labs(title    = gtitle,
                  subtitle = paste0("by ", by_cat),
                  color    = by_cat, 
                  linetype = by_cat) +
    ggplot2::xlab("Date of Reporting") +
    ggplot2::ylab(if(type == "full") {"People fully vaccinated per 100"} 
                  else {"People vaccinated with at least one dose per 100"}) +
    ggplot2::scale_x_date(labels = function(x) format(x, "%b%e,\n%Y"),
                          limits = c(min(df_c$date), max(df_c$date)),
                          breaks = "1 month") +
    ggplot2::scale_y_continuous(limits = c(0, 100), breaks = seq(0, 100, by = 20)) +
    ggplot2::scale_color_manual(values = category_color_values,
                                labels = category_labels) +
    ggplot2::scale_linetype_manual(values = category_line_values,
                                   labels = category_labels) +
    ggplot2::theme_classic() +
    ggplot2::theme(plot.title      = element_text(size  = 15, family = "Calibri", face = "bold"),
                   plot.subtitle   = element_text(size  = 14, family = "Calibri", face = "bold", margin=margin(0,0,10,0)),
                   plot.margin     = unit(c(0.5,0.5,0.5,0.5),"cm"),
                   axis.title.x    = element_text(size  = 12, family = "Calibri", margin=margin(10,0,0,0)),
                   axis.title.y    = element_text(size  = 12, family = "Calibri", margin=margin(0,10,0,0)),
                   axis.text       = element_text(size  = 10, family = "Calibri"),
                   legend.title    = element_text(size  = 12, family = "Calibri", face = "bold"),
                   legend.text     = element_text(size  = 8,  family = "Calibri"),
                   legend.position = c(0.2, 0.7))

  if(type == "full") {
    return(g + ggplot2::geom_hline(yintercept = 20, color = "black"))
  } else {
    return(g)
  }

}