index_0317.Rmd

---
title: "COVID-19 World and US View"
description: |
  Stay home, safe a life.
site: distill::distill_website
---

```{r setup, include=FALSE}
knitr::opts_chunk$set(echo = FALSE)
install.load::install_load(c('distill','dplyr', 'tidyverse', 'ggplot2', 'plotly', 'rmarkdown', 'magrittr', 'RCurl', 'lubridate', 'DT', 'ggthemes'))
```

This main page is a dashboard being built, which may be refreshed in the future. **Last updated `r format(lubridate::now("US/Pacific"), '%a, %b %d %I:%M %p')`** PST.

See "Analyses" tab for analyses performed.

## World View
```{r}
fp_data = 'https://raw.githubusercontent.com/CSSEGISandData/COVID-19/master/csse_covid_19_data/csse_covid_19_time_series/time_series_19-covid-Confirmed.csv'

# Read in data
dfc = read_csv(fp_data) %>%
  select(-Lat, -Long) %>%
  gather(date_raw, confirmed, 
         -`Province/State`, -`Country/Region`) %>%
  mutate(date = mdy(date_raw))

# Clean column names
colnames(dfc) %<>% 
  tolower() %>% 
  str_replace_all(fixed('/'), '_') %>%
  str_replace_all(fixed(' '), '_')

ggplot2::theme_set(theme_minimal())
```

```{r}
# Day since 1
df_sum <- dfc %>%
  group_by(country_region, date) %>%
  summarize(confirmed = sum(confirmed)) %>%  
  arrange(country_region, date)

df_tmp <- df_sum %>%
  filter(confirmed > 0) %>%
  mutate(day_since_start = row_number()) %>%
  ungroup() %>%
  #filter(country_region %in% c("China", 'US', 'Italy', 'France', 'Iran')) %>% #View
  highlight_key(~country_region) 
  
p <- ggplot(df_tmp, aes(x=day_since_start, 
                        y=confirmed, 
                        color=country_region)) +
  geom_line() +
  labs(title='Total Confirmed COVID-19 Cases by Days Since First Onset',
          x = 'Days Since Start of Onset',
          y = 'Total Confirmed') + 
  theme(legend.position = "none")

ggplotly(p) %>% 
  highlight('plotly_hover', 
          selectize = T, 
          # dynamic=T,
          defaultValues = c('China', 'Italy', 'US'))
```

## Countries with more than 3,000
The X-axis is days since hitting 3,000 cases.

```{r}
df_plotly <- df_sum %>%
  filter(confirmed > 3000) %>%
  mutate(days_since = row_number(),
         days_since = ifelse(country_region == 'US', days_since - 1 , days_since)) %>%
  ungroup() %>%
  highlight_key(~country_region)

p <- ggplot(df_plotly, aes(x = days_since, y = confirmed, group = country_region, text = date), color = palette_pander(7)[7]) +
  geom_line() +
  labs(title='Countries with More than 3,000 Cases',
       x = 'Days Since Hitting 3000 Confirmed Cases',
       y = 'Total Confirmed Cases',
       caption = 'Data Source: github.com/CSSEGISandData/COVID-19\ngithub.com/bryanwhiting/covid19')

ggplotly(p, tooltip = c('days_since', 'confirmed', 'country_region', 'date')) %>% 
  highlight('plotly_click', color=palette_pander(6),
            dynamic=T,
            default = 'Italy')
```

## US States Since March 1
This interactive plot shows the states that are starting to follow the trend of doubling every two days.

```{r}
# Prepare a dataframe by state. Some data processing needs to happen.
df_state <- dfc %>%
  filter(country_region == 'US',
         confirmed > 0) %>%
  group_by(province_state) %>%
  arrange(date) %>%
  mutate(days_since_first = row_number()) %>%
  arrange(province_state, date) %>%
  mutate(state2 = str_extract(province_state, '[A-Z][A-Z]$'),
         state2 = ifelse(is.na(state2), province_state, state2)) %>%
  ungroup() %>%
  select(state2, date, confirmed) %>%
  left_join(tibble(state_abb = state.abb, state2=state.name),
            by='state2') %>%
  mutate(state = ifelse(is.na(state_abb), state2, state_abb)) %>%
  select(state, date, confirmed)

# df_state %>%
#   filter(state != 'Diamond Princess') %>%
#   mutate(state = ifelse(str_detect(state, 'District|D.C.'), 'DC', state)) %>%
#   group_by(state) %>%
#   arrange(date) %>%
#   summarize(confirmed = sum(confirmed))


# Check to see how counties were aggregated
# dfc %>%
#   filter(str_detect(province_state, ', CA|California')) %>%
#   mutate(state = province_state == 'California') %>%
#   group_by(state, date) %>%
#   summarize(confirmed = sum(confirmed)) %>%
#   arrange(date, state) %>%
#   View()

df_plt = df_state %>%
  filter(date >= '2020-03-01') %>%
  ungroup() %>%
  highlight_key(~state)

p <- ggplot(df_plt, aes(x = date, y = confirmed, group=state)) +
  geom_line() + 
  scale_x_date(labels = scales::date_format("%m/%d")) +
  labs(title='Confirmed Cases in US States Since March 1',
       x = 'Date',
       y = 'Total Confirmed Cases',
       caption = 'Data Source: github.com/CSSEGISandData/COVID-19\ngithub.com/bryanwhiting/covid19')

ggplotly(p, tooltip = c('state', 'confirmed', 'date')) %>% 
  highlight('plotly_click', color=palette_pander(6),
            dynamic=T, default = 'NY')
```
> Tip: The plots are clickable. Hold shift while clicking to show multiple lines.