---
title: "Crash Data Summaries"
output:
  html_document:
    toc: true
    toc_depth: 5
    toc_float:
      collapsed: false
      smooth_scroll: true
editor_options:
  chunk_output_type: console
---


# Input Data & Configuration

## Libraries
```{r libs, eval = TRUE, echo = TRUE, results = "show", warning = FALSE, error = TRUE, message = FALSE}
date()
rm(list=ls())
library(tidyverse)
library(RColorBrewer)
library(tidycensus)
library(ggrepel)
library(leaflet)
```

## Load TOPS data
```{r loadTOPS, eval = TRUE, echo = TRUE, results = "show", warning = FALSE, error = TRUE, message = FALSE}
load(file = "data/TOPS/TOPS_data.Rda")
load(file = "data/TOPS/vuln_roles.Rda")
load(file = "data/TOPS/retrieve_date.Rda")
load(file = "data/TOPS/injury_severity.Rda")
injury_severity_pal <- colorFactor(palette = injury_severity$color, levels = injury_severity$InjSevName)

```

## Set parameters
```{r parameters, eval = TRUE, echo = TRUE, results = "show", warning = FALSE, error = TRUE, message = FALSE}
county_focus <- c("DANE")
municipality_focus <- c("MADISON")
injury_severy_focus <- c("A", "K", "B")
```

## build data summaries for city
```{r citysummaries, eval = TRUE, echo = TRUE, results = "show", warning = FALSE, error = TRUE, message = FALSE}

data_summary <- list(NULL)
# crashes by year that resulted in a pedestrian fatality or severe injury
data_summary[["crash_by_year"]] <- TOPS_data %>%
  filter(MUNINAME %in% municipality_focus) %>%
  filter(ped_inj %in% injury_severy_focus) %>%
  group_by(MUNINAME, year, vulnerable_role, ped_inj_name) %>%
  summarise(count = n_distinct(DOCTNMBR))

# crashes by race of pedestrian/bicyclist for focus year
data_summary[["crash_by_race"]] <- TOPS_data %>%
  filter(MUNINAME %in% municipality_focus) %>%
  filter(ped_inj %in% injury_severy_focus) %>%
  group_by(MUNINAME, vulnerable_role, ped_inj_name, vulnerable_race) %>%
  summarise(count = n_distinct(DOCTNMBR))

# crashes by race of driver that resulted in a pedestrian fatality or severe injury
data_summary[["crash_by_driver_race"]] <- TOPS_data %>%
  filter(MUNINAME %in% municipality_focus) %>%
  filter(ped_inj %in% injury_severy_focus) %>%
  mutate(driver_race = ifelse(ROLE1 %in% c("DR"), race_name1, ifelse(ROLE2 %in% c("DR"), race_name2, NA))) %>%
  group_by(MUNINAME, year, vulnerable_role, ped_inj_name, driver_race) %>%
  summarise(count = n_distinct(DOCTNMBR))

# crashes by age of pedestrian/bicyclist
data_summary[["crash_by_age"]] <- TOPS_data %>%
  filter(MUNINAME %in% municipality_focus) %>%
  filter(ped_inj %in% injury_severy_focus) %>%
  mutate(vulnerable_age = ifelse(ROLE1 %in% vuln_roles, age1, ifelse(ROLE2 %in% vuln_roles, age2, NA))) %>%
  group_by(MUNINAME, year, vulnerable_role, ped_inj_name, vulnerable_age) %>%
  summarise(count = n_distinct(DOCTNMBR))

# crashes by age of driver that resulted in a severe injury or fatality of a pedestrian/bicyclist
data_summary[["crash_by_driver_age"]] <- TOPS_data %>%
  filter(MUNINAME %in% municipality_focus) %>%
  filter(ped_inj %in% injury_severy_focus) %>%
  mutate(driver_age = ifelse(ROLE1 %in% c("DR"), age1, ifelse(ROLE2 %in% c("DR"), age2, NA))) %>%
  group_by(MUNINAME, year, vulnerable_role, ped_inj_name, driver_age) %>%
  summarise(count = n_distinct(DOCTNMBR))

# crashes by sex of pedestrian/bicyclist
data_summary[["crash_by_sex"]] <- TOPS_data %>%
  filter(MUNINAME %in% municipality_focus) %>%
  filter(ped_inj %in% injury_severy_focus) %>%
  mutate(vulnerable_sex = ifelse(ROLE1 %in% vuln_roles, SEX1, ifelse(ROLE2 %in% vuln_roles, SEX1, NA))) %>%
  group_by(MUNINAME, year, vulnerable_role, ped_inj_name, vulnerable_sex) %>%
  summarise(count = n_distinct(DOCTNMBR))

# crashes by sex of driver that resulted in a severe injury or fatality of a pedestrian/bicyclist
data_summary[["crash_by_driver_sex"]] <- TOPS_data %>%
  filter(MUNINAME %in% municipality_focus) %>%
  filter(ped_inj %in% injury_severy_focus) %>%
  mutate(driver_sex = ifelse(ROLE1 %in% c("DR"), SEX1, ifelse(ROLE2 %in% c("DR"), SEX2, NA))) %>%
  group_by(MUNINAME, year, vulnerable_role, ped_inj_name, driver_sex) %>%
  summarise(count = n_distinct(DOCTNMBR))

## export csv files for city ----

# make directories for city summaries
city_directory <- paste0("data_summaries/city/",municipality_focus,"/")
ifelse(!dir.exists(city_directory), dir.create(city_directory), "Folder exists already")

for(table_name in as.vector(names(data_summary[-1]))) {
  write_csv(data_summary[[table_name]], file = paste0(city_directory, municipality_focus, "-", table_name, ".csv"))
}
```


## build data summaries for county ----
```{r countysummaries, eval = TRUE, echo = TRUE, results = "show", warning = FALSE, error = TRUE, message = FALSE}

data_summary <- list(NULL)

# crashes by year that resulted in a pedestrian fatality or severe injury
data_summary[["crash_by_year"]] <- TOPS_data %>%
  filter(CNTYNAME %in% county_focus) %>%
  filter(ped_inj %in% injury_severy_focus) %>%
  group_by(CNTYNAME, year, vulnerable_role, ped_inj_name) %>%
  summarise(count = n_distinct(DOCTNMBR))

# crashes by race of pedestrian/bicyclist for focus year
data_summary[["crash_by_race"]] <- TOPS_data %>%
  filter(CNTYNAME %in% county_focus) %>%
  filter(ped_inj %in% injury_severy_focus) %>%
  group_by(CNTYNAME, vulnerable_role, ped_inj_name, vulnerable_race) %>%
  summarise(count = n_distinct(DOCTNMBR))

# crashes by race of driver that resulted in a pedestrian fatality or severe injury
data_summary[["crash_by_driver_race"]] <- TOPS_data %>%
  filter(CNTYNAME %in% county_focus) %>%
  filter(ped_inj %in% injury_severy_focus) %>%
  mutate(driver_race = ifelse(ROLE1 %in% c("DR"), race_name1, ifelse(ROLE2 %in% c("DR"), race_name2, NA))) %>%
  group_by(CNTYNAME, year, vulnerable_role, ped_inj_name, driver_race) %>%
  summarise(count = n_distinct(DOCTNMBR))

# crashes by age of pedestrian/bicyclist
data_summary[["crash_by_age"]] <- TOPS_data %>%
  filter(CNTYNAME %in% county_focus) %>%
  filter(ped_inj %in% injury_severy_focus) %>%
  mutate(vulnerable_age = ifelse(ROLE1 %in% vuln_roles, age1, ifelse(ROLE2 %in% vuln_roles, age2, NA))) %>%
  group_by(CNTYNAME, year, vulnerable_role, ped_inj_name, vulnerable_age) %>%
  summarise(count = n_distinct(DOCTNMBR))

# crashes by age of driver that resulted in a severe injury or fatality of a pedestrian/bicyclist
data_summary[["crash_by_driver_age"]] <- TOPS_data %>%
  filter(CNTYNAME %in% county_focus) %>%
  filter(ped_inj %in% injury_severy_focus) %>%
  mutate(driver_age = ifelse(ROLE1 %in% c("DR"), age1, ifelse(ROLE2 %in% c("DR"), age2, NA))) %>%
  group_by(CNTYNAME, year, vulnerable_role, ped_inj_name, driver_age) %>%
  summarise(count = n_distinct(DOCTNMBR))

# crashes by sex of pedestrian/bicyclist
data_summary[["crash_by_sex"]] <- TOPS_data %>%
  filter(CNTYNAME %in% county_focus) %>%
  filter(ped_inj %in% injury_severy_focus) %>%
  mutate(vulnerable_sex = ifelse(ROLE1 %in% vuln_roles, SEX1, ifelse(ROLE2 %in% vuln_roles, SEX1, NA))) %>%
  group_by(CNTYNAME, year, vulnerable_role, ped_inj_name, vulnerable_sex) %>%
  summarise(count = n_distinct(DOCTNMBR))

# crashes by sex of driver that resulted in a severe injury or fatality of a pedestrian/bicyclist
data_summary[["crash_by_driver_sex"]] <- TOPS_data %>%
  filter(CNTYNAME %in% county_focus) %>%
  filter(ped_inj %in% injury_severy_focus) %>%
  mutate(driver_sex = ifelse(ROLE1 %in% c("DR"), SEX1, ifelse(ROLE2 %in% c("DR"), SEX2, NA))) %>%
  group_by(CNTYNAME, year, vulnerable_role, ped_inj_name, driver_sex) %>%
  summarise(count = n_distinct(DOCTNMBR))

## export csv files for county ----

# make directories for county summaries

county_directory <- paste0("data_summaries/county/",county_focus,"/")
ifelse(!dir.exists(county_directory), dir.create(county_directory), "Folder exists already")


for(table_name in as.vector(names(data_summary[-1]))) {
  write_csv(data_summary[[table_name]], file = paste0("data_summaries/county/", county_focus, "/", county_focus, "_", table_name, ".csv"))
}
```


## Make figures for county summaries
```{r countysummaryfigures, eval = TRUE, echo = TRUE, results = "show", warning = FALSE, error = TRUE, message = FALSE}
data_summary[["crash_by_age"]] %>%
  filter(vulnerable_age < 18,
         year != 2025) %>%
  group_by(year, ped_inj_name) %>%
  summarise(count = sum(count)) %>%
  ggplot() +
  geom_col(aes(x = year,
               y = count,
               fill = ped_inj_name)) +
  scale_fill_manual(values = injury_severity_pal(injury_severity %>% filter(code %in% injury_severy_focus) %>% pull(InjSevName))) +
  labs(title = paste0("Pedestrian & bicyclists under 18yo killed or injured in ", str_to_title(county_focus), " county"),
       x = "Year",
       y = "Total crashes",
       fill = "Injury Severity",
       caption = paste0("crash data from UW TOPS lab - retrieved ",
                        strftime(retrieve_date, format = "%m/%Y"),
                        " per direction of the WisDOT Bureau of Transportation Safety")) +
  theme(plot.caption = element_text(color = "grey"))
ggsave(file = paste0(county_directory, county_focus, "_youth_crashes.pdf"),
       height = 8.5,
       width = 11,
       units = "in")

TOPS_data %>%
  filter(CNTYNAME %in% county_focus) %>%
  filter(year == 2024) %>%
  filter(ped_inj %in% "K") %>%
  mutate(vulnerable_age = ifelse(ROLE1 %in% vuln_roles, age1, ifelse(ROLE2 %in% vuln_roles, age2, NA))) %>%
  filter(vulnerable_age < 18)

```

## Make figures for statewide summaries
```{r statesummary, eval = TRUE, echo = TRUE, results = "show", warning = FALSE, error = TRUE, message = FALSE}
TOPS_data %>%
  filter(!is.na(vulnerable_role)) %>%
  filter(vulnerable_role == "Pedestrian") %>%
  mutate(driver_vehtype = ifelse(ROLE1 %in% c("DR"), VEHTYPE1, ifelse(ROLE2 %in% c("DR"), VEHTYPE2, NA))) %>%
  filter(driver_vehtype %in% c("CAR", "SUV", "UT TRK")) %>%
  group_by(vulnerable_role, ped_inj_name, driver_vehtype) %>%
  summarise(count = n_distinct(DOCTNMBR)) %>%
  pivot_wider(names_from = driver_vehtype, values_from = count) %>%
  ungroup() %>%
  mutate(total = (CAR + SUV + `UT TRK`)) %>%
  mutate(car_percent = CAR/total * 100,
         suv_percent = SUV/total * 100,
         pickup_percent = `UT TRK`/total * 100) %>%
  select(vulnerable_role, ped_inj_name, car_percent, suv_percent, pickup_percent) %>%
  mutate(Car = car_percent, SUV = suv_percent, "Pickup Truck" = pickup_percent) %>%
  pivot_longer(cols = c(Car, SUV, "Pickup Truck")) %>%
  ggplot(aes(x = ped_inj_name,
               y = value,
               fill = name)) +
  geom_col(position = position_stack()) +
  geom_text(aes(label = paste0(round(value, 1), "%")),
            position = position_stack(vjust = 0.5)) +
  theme(axis.text.x = element_text(angle = 30, hjust = 1),
        plot.caption = element_text(color = "grey")) +
  scale_y_continuous(expand = expansion(mult = c(0, 0))) +
  labs(title = "Impact of type of vehicle on severity of injury",
       subtitle = paste0("Crashes statewide - Wisconsin - 2017-", year(max(TOPS_data$date))),
       y = "Percent of crashes",
       x = "Severity of injury",
       fill = "Type of vehicle",
       caption = paste0("crash data from UW TOPS lab - retrieved ",
                        strftime(retrieve_date, format = "%m/%Y"),
                        " per direction of the WisDOT Bureau of Transportation Safety"))
ggsave(file = paste0("figures/crash_summaries/state/", "VehType_InjSev.pdf"),
       height = 8.5,
       width = 11,
       units = "in")


TOPS_data %>%
  filter(!is.na(vulnerable_role)) %>%
  filter(vulnerable_role == "Pedestrian") %>%
  filter(ped_inj %in% injury_severy_focus) %>%
  mutate(driver_vehtype = ifelse(ROLE1 %in% c("DR"), VEHTYPE1, ifelse(ROLE2 %in% c("DR"), VEHTYPE2, NA))) %>%
  filter(driver_vehtype %in% c("CAR", "SUV", "UT TRK")) %>%
  group_by(vulnerable_role, driver_vehtype, year) %>%
  summarise(count = n_distinct(DOCTNMBR)) %>%
  pivot_wider(names_from = driver_vehtype, values_from = count) %>%
  ungroup() %>%
  mutate(total = (CAR + SUV + `UT TRK`)) %>%
  mutate(car_percent = CAR/total * 100,
         suv_percent = SUV/total * 100,
         pickup_percent = `UT TRK`/total * 100) %>%
  select(year, vulnerable_role, car_percent, suv_percent, pickup_percent) %>%
  pivot_longer(cols = c(car_percent, suv_percent, pickup_percent)) %>%
  ggplot() +
  geom_line(aes(x = as.double(as.character(year)),
               y = value,
               color = name))

TOPS_data %>%
  filter(!is.na(vulnerable_role)) %>%
  filter(vulnerable_role == "Pedestrian") %>%
  filter(ped_inj %in% "K") %>%
  mutate(driver_vehtype = ifelse(ROLE1 %in% c("DR"), VEHTYPE1, ifelse(ROLE2 %in% c("DR"), VEHTYPE2, NA))) %>%
  filter(driver_vehtype %in% c("CAR", "SUV", "UT TRK")) %>%
  group_by(vulnerable_role, driver_vehtype, year) %>%
  summarise(count = n_distinct(DOCTNMBR)) %>%
  pivot_wider(names_from = driver_vehtype, values_from = count) %>%
  ungroup() %>%
  mutate(total = (CAR + SUV + `UT TRK`)) %>%
  mutate(car_percent = CAR/total * 100,
         suv_percent = SUV/total * 100,
         pickup_percent = `UT TRK`/total * 100) %>%
  select(year, vulnerable_role, car_percent, suv_percent, pickup_percent) %>%
  pivot_longer(cols = c(car_percent, suv_percent, pickup_percent)) %>%
  ggplot() +
  geom_line(aes(x = as.double(as.character(year)),
               y = value,
               color = name))

```