Exploratory Data Analysis.Rmd

---
title: "tdor"
output: html_notebook
author: Dave Parr & Mary Cleaton
---

## How many deaths have occured over time by country?

```{r, fig.dim=c(10,10)}
library(tdor)
library(tidyverse)
tdor <- tdor

tdor %>%
  group_by(Year, Country) %>%
  summarise(DeathCount = n()) %>%
  ggplot(aes(x = Year, y = DeathCount)) +
  geom_point() +
  facet_wrap( ~ Country, scales = "free_y")
```

* Brazil has a much higher recorded deaths rate than any other country. 
* Some countries rates and/or reporting rates are increasing
* Other countries rates and/or reporting rates show no clear pattern over time

## Gun Deaths

USA has a very high volume of reported deaths. Is this potentially related to gun availability?

```{r}
tdor %>%
  group_by(Country) %>%
  summarise(TotalDeaths = n()) -> country_totals

tdor %>%
  mutate(
    Shot = case_when(
      str_detect(`Cause of death`, "shot") ~ TRUE, 
      TRUE ~ FALSE)
    ) %>%
  group_by(Country, Shot) %>%
  summarise(DeathCount = n()) %>%
  left_join(country_totals) %>%
  filter(TotalDeaths > 2 & Shot == TRUE) %>%
  mutate(DeathPrct = (DeathCount / TotalDeaths) * 100) -> tdorShotPercent

ggplot(data = tdorShotPercent, aes(
  x = reorder(Country,-DeathPrct),
  y = DeathPrct,
  fill = Shot
)) +
  geom_col() +
  theme_minimal() +
  theme(axis.text.x = element_text(
    angle = 90,
    hjust = 1,
    vjust = 0.5
  )) +
  labs(title = "Percentage of Deaths by shooting",
       subtitle = "in countries with more than 2 Deaths",
       y = "Deaths (%)")
```


* Gun related deaths are very variable between countries
* Central and South American countries plus USA appear to have the highest proportions of gun deaths