---
title: "COVID-19"
output:
  flexdashboard::flex_dashboard:
    logo: rbmv_curve.png
    orientation: rows
    source_code: embed
    theme: flatly
---



```{r setup, include=FALSE}
library(flexdashboard)
```

```{r, include=FALSE}
# Packages
library(tidyverse)
library(lubridate)
library(rvest)
library(sf)
library(leaflet)
library(plotly)
library(crosstalk)
library(rbmv)
library(classInt)

# imported functions
`%<>%` <- magrittr::`%<>%`

# adjust depending on when kniting
page_updated <- today()

# create the friendly label for value box
page_updated_string <- str_glue("{wday(page_updated, label = T)} {month(page_updated, label = T)} {day(page_updated)}, {year(page_updated)}")

# custom functions
get_bins <- function(.data, bins = 8) {
  #' use kmeans to automatically set sensible bins that represent the data well
  #' requires 'classInt' package
  #' @return a vector with 8 default bins
  intervals <- classIntervals(.data, n = bins, style = "kmeans", rtimes = 5)
  return(round(intervals$brks))
}

# API key
mapbox <- read_lines(".mapbox-key")

# grab the html from the dph website
dph_covid19_page <- read_html("http://www.publichealth.lacounty.gov/media/Coronavirus/locations.htm")

# only extract the table tags
dph_covid19_tbls <- dph_covid19_page %>%
  html_nodes("table") %>%
  html_table()

# just focu on grabbing the locations, cases, and deaths for lb and pass
dph_covid19_messy <- dph_covid19_tbls[[1]] %>%
  repair_names() %>%
  janitor::clean_names() %>%
  rename(locations = x1, total_cases = x2) %>%
  select(-x3:-x5)

# what we want to keep
keep <- "Long Beach|Pasadena"

# add the labels so we can 'tidyr::spread' in preparation for a future join
label <- c("cases_final", "cases_final", "deaths_final", "deaths_final")

# save only the lb/pass data
lb_pass <- dph_covid19_messy %>%
  filter(str_detect(locations, keep)) %>%
  slice(1:4)

# add the labels
lb_pass$label <- label

# now spread and ensure the data types are correct for future join
lb_pass %<>%
  spread(label, total_cases) %>%
  mutate(
    locations = str_replace(locations, "--", "City of"),
    cases_final = parse_number(cases_final),
    deaths_final = parse_number(deaths_final))

# read the data from the DPH shiny app
data <- read_csv("https://lacdph.shinyapps.io/covid19_surveillance_dashboard/_w_6c4098c3/session/c8a099ca7c06ead4f764f6803e9e75be/download/download2?w=6c4098c3")

# rename and join
dph_covid19 <- data %>%
  select(-X1) %>%
  rename(locations = geo_merge) %>%
  full_join(
    .,
    lb_pass,
    by = c("locations", "cases_final", "deaths_final"))

lac_total_cases <- dph_covid19 %>%
  summarise(total_cases = sum(cases_final))

lac_total_deaths <- dph_covid19 %>%
  summarise(total_deaths = sum(deaths_final))

# read the gis data
gis <- st_read(
  dsn = "data/gis/geo_export_871186bb-3266-4a1c-a0f6-8cda86b55d55.shp",
  layer = "geo_export_871186bb-3266-4a1c-a0f6-8cda86b55d55",
  quiet = T
)

# make sure the st is set correctly
gis %<>% st_transform(crs = 4326) %>%
  mutate(label = as.character.factor(label))

# join to cleaned data
lac_covid19 <- left_join(
  dph_covid19,
  gis,
  by = c("locations" = "label")) %>%
  rename(total_cases = cases_final, total_deaths = deaths_final) %>%
  st_as_sf()

# get time series data from GitHub
covid19_cases <- read_csv("https://raw.githubusercontent.com/CSSEGISandData/COVID-19/master/csse_covid_19_data/csse_covid_19_time_series/time_series_covid19_confirmed_US.csv")

# same as above
covid19_deaths <- read_csv("https://raw.githubusercontent.com/CSSEGISandData/COVID-19/master/csse_covid_19_data/csse_covid_19_time_series/time_series_covid19_deaths_US.csv")

# function to make sure we're getting the most current world data
make_link <- function(date) {
  date_format <- format(date, format = "%m-%d-%Y")
  link <- str_glue("https://raw.githubusercontent.com/CSSEGISandData/COVID-19/master/csse_covid_19_data/csse_covid_19_daily_reports/{date_format}.csv")
  return(link)
}

# test the link
test_data <- function(date) {
  link <- make_link(date)
  status <- httr::http_status(httr::GET(link))
  return(status$category)
}

# if the link is correct, then get the world data
get_world_data <- function(date) {
  if (test_data(date) == "Success") {
    return(make_link(date))
  } else if (test_data(date) == "Client error") {
    return(make_link(date - 1))
  }
}

# create link based on today's date
link <- get_world_data(today())

# now read it!
world <- read_csv(link)
world %<>% rename(long = Long_, lat = Lat)

covid19_time_series_cases <- covid19_cases %>%
  filter(Admin2 == "Los Angeles") %>%
  gather("date", "cases", 12:ncol(covid19_cases)) %>%
  select(Admin2, date, cases) %>%
  rename(county = Admin2) %>%
  mutate(date = parse_date(date, "%m/%d/%y"),
         case_type = "New Confirmed Case") %>%
  distinct()

covid19_time_series_deaths <- covid19_deaths %>%
  filter(Admin2 == "Los Angeles") %>%
  gather("date", "cases", 13:ncol(covid19_deaths)) %>%
  select(Admin2, date, cases) %>%
  rename(county = Admin2) %>%
  mutate(date = parse_date(date, "%m/%d/%y"),
         case_type = "Death") %>%
  distinct()

covid19_time_series <- bind_rows(covid19_time_series_cases, covid19_time_series_deaths)

covid19_cases %<>% mutate(Combined_Key = str_remove(Combined_Key, ", US"))

covid19_cases_top10_us <- covid19_cases %>%
  arrange(desc(.[[ncol(covid19_cases)]])) %>%
  slice(1:10) %>%
  gather("date", "cases", 12:ncol(covid19_cases)) %>%
  select(Combined_Key, date, cases) %>%
  rename(location = Combined_Key) %>%
  mutate(
    location = str_remove(location, ", US"),
    date = parse_date(date, "%m/%d/%y"),
    case_type = "New Confirmed Case") %>%
  distinct()

covid19_state_cases <- covid19_cases %>%
  filter(iso3 == "USA") %>%
  gather("date", "cases", 12:ncol(covid19_cases)) %>%
  rename(state = Province_State) %>%
  group_by(state, date) %>%
  summarise(cases = sum(cases)) %>%
  mutate(date = parse_date(date, "%m/%d/%y"))

light_blue <- rbmv_pal("main", plotly = T)[1]
ruby <- rbmv_pal("main", plotly = T)[2]
pale_black <- rbmv_pal("main", plotly = T)[5]

sd <- SharedData$new(covid19_cases_top10_us, ~location, "Select a city")
```

LA County {data-icon="fa-map"}
=====================================

Row {data-height=510}
-------------------------------------

###

```{r}
bins <- get_bins(lac_covid19$total_cases, bins = 12)

pal <- colorBin(
  palette = rbmv_pal("spectrum", plotly = T),
  domain = lac_covid19$total_cases,
  bins = bins)

lac_covid19 %>%
  leaflet() %>%
  addTiles(mapbox, attribution = paste("Data from Los Angeles County Department of Public Health COVID-19 location table")) %>%
  addPolygons(
    fillColor = ~pal(total_cases),
    weight = 1.5,
    fillOpacity = 0.7,
    smoothFactor = 0.5,
    color = "white",
    label = ~str_glue(
      "{locations}

      Total cases: {scales::comma(total_cases)}

      Total deaths: {scales::comma(total_deaths)}

      Population: {scales::comma(population)}
") %>%
      map(htmltools::HTML),
    labelOptions = labelOptions(
      style = list("font-weight" = "normal", padding = "3px 8px"),
      textsize = "15px",
      direction = "auto")) %>%
  addLegend(
    pal = pal, values = ~total_cases, opacity = 0.7,
    title = NULL, position = "bottomright") %>%
  setView(-118.2, 34, zoom = 9.5)
```

Row {data-height=400}
-------------------------------------

###

```{r}
source <- list(
  x = 1, y = -0.25, text = "Source: 2019 Novel Coronavirus COVID-19 (2019-nCoV)\nData Repository by Johns Hopkins CSSE",
  showarrow = F, xref = "paper", yref = "paper", xanchor = "right", yanchor = "auto",
  xshif = 0, yshift = 0, font = list(size = 12, color = "grey")
)

base <- plot_ly(sd, color = I(rbmv_pal(plotly = T)[6]), height = 320) %>%
  group_by(location)

left_bar <- base %>%
  add_bars(x = ~cases, y = ~location, hoverinfo = "text",
           text = ~paste("Click bar to highlight line
on the left")) %>%
  layout(
    barmode = "overlay",
    xaxis = list(title = ""),
    yaxis = list(
      title = "",
      categoryorder = "max ascending")
  )

right_line <- base %>%
  add_lines(x = ~date, y = ~cases, alpha = 0.3, hoverinfo = "text",
          text = ~paste("
 Location: ", location,
                        "
 Case type:  ", case_type, "x Log Scale",
                        "
 Cases: ", scales::comma(cases))) %>%
  layout(xaxis = list(title = ""),
         yaxis = list(type = "log"))

subplot(left_bar, right_line, titleX = TRUE, widths = c(.3, .7)) %>%
  layout(title = "How does Los Angeles County compare to other major outbreak centers",
         margin = list(l = 120, b = 40), annotations = source) %>%
  hide_legend() %>%
  highlight("plotly_click", color = rbmv_pal(plotly = T)[2]) %>%
  config(displayModeBar = FALSE)
```

Row {data-height=110}
-------------------------------------

###

```{r}
valueBox(page_updated_string, color = light_blue, icon = "fa-calendar", caption = "Data pulled from both LAC DPH and John Hopkins CSSE")
```


###

```{r}
total_cases <- scales::comma(last(lac_total_cases$total_cases))
valueBox(total_cases, color = ruby, icon = "fa-ambulance", caption = "Total Lab Confirmed COVID-19 Cases in LAC")
```

###

```{r}
total_deaths <- scales::comma(last(lac_total_deaths$total_deaths))
valueBox(total_deaths, color = pale_black, icon = "fa-medkit", caption = "Total COVID-19 Related Deaths")
```

US States {data-icon="fa-chart-line"}
=====================================

###

```{r}
bins <- get_bins(covid19_state_cases$cases, bins = 8)
colors <- RColorBrewer::brewer.pal(9, "YlOrRd")

normalize <- function(x) round((x-min(x))/(max(x)-min(x)), 4)

state_colors <- tibble(range = bins, hex = colors) %>%
  mutate(range = normalize(range))

plot_ly(covid19_state_cases) %>%
  filter(date >= "2020-03-15") %>%
  add_heatmap(x = ~date, y = ~state, z = ~cases,
              colorscale = state_colors, xgap = .8,
              opacity = .85, showscale = F,
              text = ~paste(
                "Date: ", date, "
State: ", state,
                "
Total confirmed cases: ", scales::comma(cases)),
              hoverinfo = "text") %>%
  layout(
    title = "COVID19 Cases by State",
    xaxis = list(title = ""),
    yaxis = list(title = "", categoryorder = "max ascending", side = "right")) %>%
  config(displayModeBar = FALSE)
```


Global {data-icon="fa-globe"}
=====================================

###

```{r}
bins <- get_bins(world$Confirmed, bins = 12)

pal <- colorBin(
  palette = rbmv_pal("warm", plotly = T),
  domain = world$Confirmed,
  bins = bins)

label <- str_glue("Location: {world$Combined_Key}
      
Cases: {scales::comma(world$Confirmed)}
      
Deaths: {scales::comma(world$Deaths)}
      
Active: {scales::comma(world$Active)}
      
Recovered: {scales::comma(world$Recovered)}
      
Last Update: {world$Last_Update}") %>%
  lapply(htmltools::HTML)

world %>%
  leaflet() %>%
  addTiles(mapbox, attribution = paste("Data from Novel Coronavirus (COVID-19) Cases, provided by JHU CSSE")) %>%
  addCircleMarkers(
    lng = ~long, lat = ~lat,
    label = ~label,
    color = ~pal(Confirmed),
    radius = ~ifelse(Confirmed >= 30000, 25, 15),
    stroke = FALSE, fillOpacity = .7,
    clusterOptions = markerClusterOptions(
      showCoverageOnHover = FALSE),
  labelOptions = labelOptions(
    style = list("font-weight" = "normal", padding = "3px 8px"),
    textsize = "15px",
    direction = "auto")) %>%
  addLegend(title = "COVID-19 Cases", opacity = 0.6, pal = pal, values = ~Confirmed, position = "bottomright")
```