by_county_html.Rmd

---
output: pdf_document
---

```{r setup, include=FALSE}
knitr::opts_chunk$set(fig.align = "center", fig.width = 7.5, fig.height=5, echo = FALSE, message=FALSE, warning=FALSE)
# knitr::opts_chunk$set(fig.align = "center", fig.width = 8, fig.height=5, echo = FALSE, message=FALSE, warning=FALSE)
```

```{r load data}
library(readr)
library(tidyverse)
library(readxl)
library(zoo)
library(ggthemes)
library(extrafont)
library(patchwork)
library(pander)

lenged_font_size <- 7

loadfonts(device = "win")


cc <- read_csv("data_world_download/COVID-19 Activity.csv",
col_types = cols(REPORT_DATE = col_date(format = "%Y-%m-%d"),
COUNTY_NAME = col_character(), COUNTY_FIPS_NUMBER = col_character(),
PEOPLE_POSITIVE_CASES_COUNT = col_number(),
PEOPLE_POSITIVE_NEW_CASES_COUNT = col_number(),
PEOPLE_DEATH_COUNT = col_number(),
PEOPLE_DEATH_NEW_COUNT = col_number()))

# cc <- read_csv("data_world_download/COVID-19 Activity.csv",
#     col_types = cols(county_name = col_character(),
#         county_fips_number = col_character(),
#         people_positive_cases_count = col_number(),
#         report_date = col_date(format = "%Y-%m-%d"),
#         people_death_new_count = col_number(),
#         people_positive_new_cases_count = col_number(),
#         people_death_count = col_number()))



date_of_data <- format(max(cc$report_date, "%d-%b-%Y"))
caption1 <- paste("Data updated on ", date_of_data, sep = "") 
```

```{r run code}

# Create pallet and color look up -----------------------------------------
soi <- c("Brunswick, North Carolina, US",
         "Richland, South Carolina, US",
         "Clark, Ohio, US",
         "Licking, Ohio, US",
         "Durham, North Carolina, US",
         "Guilford, North Carolina, US",
         "Chesterfield, Virginia, US",
         "Greenville, South Carolina, US",
         "Charleston, South Carolina, US", 
         "Horry, South Carolina, US")

# soi <- c("Los Angeles, California", "Wake, North Carolina")


soi <- sort(soi)

n_of_soi <- length(soi)
pallet <-   paletteer::paletteer_c("pals::kovesi.cyclic_mygbm_30_95_c78_s25", n = n_of_soi)
pallet <- as.character(pallet)

soi <- data.frame(soi, pallet, stringsAsFactors = FALSE)



# create final table to run charts ----------------------------------------


pop_lookup <- read_rds(here::here("population_lookup.rds"))


soi_table <- cc %>%
  select(Date = REPORT_DATE,
         Admin2 = COUNTY_NAME,
         Province_State = PROVINCE_STATE_NAME,
         COUNTRY_SHORT_NAME,
         country_code = COUNTRY_ALPHA_2_CODE,
         CONTINENT_NAME,
         Cases = PEOPLE_POSITIVE_CASES_COUNT,
         Difference = PEOPLE_POSITIVE_NEW_CASES_COUNT,
         Deaths = PEOPLE_DEATH_NEW_COUNT,
         Cumulative_Deaths = PEOPLE_DEATH_COUNT,
         ) %>%
  mutate(Combined_Key = paste(Admin2, Province_State, country_code, sep = ", ")) %>%
  left_join(pop_lookup, by = "Combined_Key") %>%
  filter(Combined_Key %in% soi$soi)  %>%
  mutate(infection_Rate = (Cases/Population_Count)*100000)

# soi_table <- cc %>%
#   select(Date = report_date,
#          Admin2 = county_name,
#          Province_State = province_state_name,
#          country_short_name,
#          country_code = country_alpha_2_code,
#          continent_name,
#          Cases = people_positive_cases_count,
#          Difference = people_positive_new_cases_count,
#          Deaths = people_death_new_count,
#          Cumulative_Deaths = people_death_count,
#          ) %>%
#   mutate(Combined_Key = paste(Admin2, Province_State, country_code, sep = ", ")) %>%
#   left_join(pop_lookup, by = "Combined_Key") %>%
#   filter(Combined_Key %in% soi$soi)  %>%
#   mutate(infection_Rate = (Cases/Population_Count)*100000)






# create charts for report -----------------------------------------------------------


# PLOT MOVING AVERAGE OF NEW CASES
#create color table
soi_per_new_colors <- soi_table %>% 
  left_join(soi, by = c("Combined_Key" = "soi")) %>%  
  distinct(Combined_Key, pallet) %>%  
  arrange(Combined_Key)

jColors <- soi_per_new_colors$pallet
names(jColors) <- soi_per_new_colors$Combined_Key


rm(pallet, n_of_soi)

tmax <- max(soi_table$Difference, na.rm = TRUE) - (max(soi_table$Difference, na.rm = TRUE)*.7)
  
#create chart for moving average
soi_per_new <- soi_table %>% 
  arrange(Combined_Key) %>% 
  left_join(soi, by = c("Combined_Key" = "soi")) %>% 
  ggplot(aes(x = Date, y = Difference)) + 
  geom_smooth(aes(colour = Combined_Key ), size = 1.25, se = FALSE) +
  scale_colour_manual(values =  jColors) +
  theme_clean() +
  theme(legend.position = "bottom",
        text = element_text(family = "Times New Roman"),
        legend.text=element_text(size= lenged_font_size)) + 
  labs(title = "Daily New Cases -  \"loess\" smoothing", 
       subtitle = "LOESS = Local weighted regression",
       y = "Number of cases per day", 
       colour = "") + 
  facet_wrap(~Province_State, nrow = 2)

soi_per_new_actual <- soi_table %>% 
  arrange(Combined_Key) %>% 
  left_join(soi, by = c("Combined_Key" = "soi")) %>% 
  ggplot(aes(x = Date, y = Difference)) + 
  geom_line(aes(x = Date, y = Difference, colour = Combined_Key)) +
  scale_colour_manual(values =  jColors) +
  ylim(0,800) +
  theme_clean() +
  theme(legend.position = "bottom",
        text = element_text(family = "Times New Roman"),
        legend.text=element_text(size= lenged_font_size)) + 
  labs(title = "Daily New Cases", 
       y = "Number of cases per day", 
       caption = "*Dotted line is actual new cases per day",
       colour = "") + 
  facet_wrap(~Province_State, nrow = 2)





# CUMULATIVE CASES
#create color table    
soi_per_total_colors <- soi_table %>% 
  left_join(soi, by = c("Combined_Key" = "soi")) %>%
  distinct(Combined_Key, pallet) %>%  
  arrange(Combined_Key)

jColors2 <- soi_per_total_colors$pallet
names(jColors2) <- soi_per_total_colors$Combined_Key

soi_per_total <- soi_table %>% 
  arrange(Combined_Key) %>% 
  ggplot() + 
  geom_col(aes(x = Date, y = Cases, fill = Combined_Key)) +
  scale_fill_manual(values =  jColors2) +
  theme_clean() +
  theme(legend.position = "none",
        text = element_text(family = "Times New Roman")) + 
  labs(title = "Cumulative Cases", 
       y = "", 
       fill = "",
       caption = caption1 )  


# PLOT infection rate
#create color table
soi_per_ir_colors <- soi_table %>% 
  left_join(soi, by = c("Combined_Key" = "soi")) %>%  
  distinct(Combined_Key, pallet) %>%  
  arrange(Combined_Key)

jColors_ir <- soi_per_ir_colors$pallet
names(jColors_ir) <- soi_per_ir_colors$Combined_Key


#create chart infection rates
soi_per_ir <- soi_table %>% 
  arrange(Combined_Key) %>% 
  left_join(soi, by = c("Combined_Key" = "soi")) %>% 
  ggplot(aes(x = Date, y = infection_Rate)) + 
  geom_line(aes(colour = Combined_Key), size = 1.25) + 
  scale_colour_manual(values =  jColors_ir) +
  theme_clean() +
  theme(legend.position = "bottom",
        text = element_text(family = "Times New Roman"),
        legend.text=element_text(size= lenged_font_size)) + 
  labs(title = "Number of Cumulative Cases per 100,000 people", 
       y = "Number of cases per day", 
       colour = "") + 
  facet_wrap(~Province_State, nrow = 2)



# running fatalities
#create color table    
soi_per_deaths_colors <- soi_table %>% 
  left_join(soi, by = c("Combined_Key" = "soi")) %>%  
  distinct(Combined_Key, pallet) %>%  
  arrange(Combined_Key)

jColors3 <- soi_per_deaths_colors$pallet
names(jColors3) <- soi_per_deaths_colors$Combined_Key

#create color chart  
soi_per_new_deaths <- soi_table %>% 
  arrange(Combined_Key) %>% 
  ggplot(aes(x = Date, y = Deaths, colour = Combined_Key)) + 
  geom_smooth(se = FALSE) + 
  scale_colour_manual(values =  jColors3) +
  theme_clean() +
  theme(legend.position = "none",
        text = element_text(family = "Times New Roman"),
        legend.text=element_text(size=5)) + 
  labs(title = "Daily Fatalities LOESS smoothing", 
       y = "Number per day", 
       colour = "",
       caption = caption1)  +
  facet_wrap(~Province_State, nrow = 2)

soi_per_new_deaths_acctual <- soi_table %>% 
  arrange(Combined_Key) %>% 
  ggplot(aes(x = Date, y = Deaths, colour = Combined_Key)) + 
  geom_line() + 
  scale_colour_manual(values =  jColors3) +
  theme_clean() +
  theme(legend.position = "none",
        text = element_text(family = "Times New Roman")) + 
  labs(title = "Daily Fatalities -actual", 
       y = "Number per day", 
       colour = "",
       caption = caption1)  +
  facet_wrap(~Combined_Key, ncol = 3)

# CUMULATIVE fatalities
#create color table    
soi_per_total_colors <- soi_table %>% 
  left_join(soi, by = c("Combined_Key" = "soi")) %>%
  distinct(Combined_Key, pallet) %>%  
  arrange(Combined_Key)

jColors4 <- soi_per_total_colors$pallet
names(jColors4) <- soi_per_total_colors$Combined_Key

soi_per_total_deaths <- soi_table %>% 
  arrange(Combined_Key) %>% 
  ggplot() + 
  geom_col(aes(x = Date, y = Cumulative_Deaths, fill = Combined_Key)) +
  scale_fill_manual(values =  jColors2) +
  theme_clean() +
  theme(legend.position = "none",
        text = element_text(family = "Times New Roman")) + 
  labs(title = "Cumulative Fatalities", 
       y = "", 
       fill = "",
       caption = caption1 )  



population_table <- soi_table %>%  
  distinct(Combined_Key, population = Population_Count)  %>%  
  rename(Location = Combined_Key) %>% 
  arrange(desc(population)) %>% 
  mutate(population = scales::comma(population)) %>% 
  gt::gt()
  
```
  
    
## Counties of Interest  
Data from New York Times and Census data

`r caption1`  

```{r plot}
soi_per_new 
soi_per_new_actual
soi_per_ir
population_table
soi_per_new_deaths
soi_per_new_deaths_acctual
soi_per_total + soi_per_total_deaths

```

```{r places with change}
#create parameters for dataes
t2_end <- Sys.Date() -1
t2_start <- t2_end - 7

t1_end <- t2_end - 14
t1_start <- t2_start -14


t1_days <- seq(t1_start, t1_end, 1)
t2_days <- seq(t2_start, t2_end, 1)

#create function to find percentage
pec_chng <- function(t1, t2) { 
      (t2-t1)/t1 }




jColors10 <- soi_per_total_colors$pallet
names(jColors10) <- soi_per_total_colors$Combined_Key


#create data table for calculation
soi_table %>%  
  filter(Date > "2020-03-01") %>%  
  group_by(Combined_Key) %>% 
  filter(Date %in% t1_days | Date %in% t2_days) %>% 
  mutate(Test_Period = ifelse(Date %in% t1_days, "t1", "t2")) %>%  
  ungroup() %>%  
  group_by(Combined_Key, Test_Period) %>% 
  summarize(weekly_average = mean(Difference)) %>%  
  spread(Test_Period, value = weekly_average) %>%  
  filter(t2 > 0) %>% 
  filter(t1 > 0) %>% 
  mutate(percentage_change = pec_chng(t1,t2)) %>%  
  filter(Combined_Key %in% soi$soi) %>% 
  ungroup() %>% 
  mutate(Combined_Key = as.factor(Combined_Key)) %>%  
  mutate(Combined_Key = fct_reorder(Combined_Key, percentage_change)) %>%  
  ggplot(aes(x = Combined_Key, y = percentage_change, fill = Combined_Key)) +
  geom_col() +
  scale_fill_manual(values =  jColors10) +
  coord_flip() + 
  theme_clean() +
  theme(legend.position = "none",
        text = element_text(family = "Times New Roman")) + 
  labs(title = "Percentage Change in Daily", 
       subtitle = "Calculated by comparing the weekly average for this week and a week 14 days ago",
       y = "", 
       fill = "",
       caption = caption1 )  
  

  





```