trace_data.Rmd

---
title: "The Trace Data Cleaning and Graphing"
author: "Yao Yu"
date: "11/19/2019"
output: html_document
---

```{r setup, include=FALSE}
knitr::opts_chunk$set(echo = FALSE)
library(ggmap)
library(gganimate)
library(ggthemes)
library(tidyverse)
```

```{r The Trace Okland Data Cleaning}

# Reading in the trace dataset downloaded from here: https://www.thetrace.org/violent-crime-data/

trace_OK <- read.csv("the_trace_data/Oakland_CA/PRR# 21044.csv")

# Cleaning the data to add a year column. first_char is used to detec the first
# character in the address to see if the address is valid (starts with a
# number). Uses as.numeric to check if the first character is a number or not
# and then drops all NAs (non numeric characters).

# I used supressWarnings after finding this StackOverflow:
# https://stackoverflow.com/questions/14984989/how-to-avoid-warning-when-introducing-nas-by-coercion

trace_OK <- trace_OK %>%
  mutate(year = format(as.Date(trace_OK$Date, format="%d-%b-%Y"),"%Y")) %>%
  mutate(first_chr = substr(trace_OK$Location..Address, 1, 1)) %>%
  mutate(number = suppressWarnings(as.numeric(first_chr))) %>%
  drop_na(number)

# Filters data for violent crimes that involve a firearm to look specifically at
# gun violence.

trace_OK <- trace_OK %>%
  filter(Weapon == "Firearm") %>%
  mutate(year = as.integer(year)) %>%
  mutate(year = ifelse(year == 10, 2010, year),
         year = ifelse(year == 11, 2011, year),
         year = ifelse(year == 12, 2012, year),
         year = ifelse(year == 13, 2013, year),
         year = ifelse(year == 14, 2014, year),
         year = ifelse(year == 15, 2015, year),
         year = ifelse(year == 16, 2016, year),
         year = ifelse(year == 17, 2017, year))

# Writes clean data to a new csv file

write.csv(x = trace_OK, file = "the_trace_data/Oakland_CA/OK_trace_cleaned.csv")
```

```{r Graphing Trace Okland data}

# This following method of graphing points on a map can be found here:
# https://www.littlemissdata.com/blog/maps

# Reads in clean csv data with a specification between 2011-2017 to stay
# consistent with the violent crimes data

trace_OK <- read.csv("clean-data/trace_OK.csv") %>%
  filter(year %in% c(2011, 2012, 2013, 2014, 2015, 2016, 2017))

#Google api key (hidden for obvious reasons)

ggmap::register_google(key = "***************************************")

# Gets a google map image of Oakland using Google Map's api

p <- ggmap(get_googlemap(center = c(lon = -122.225, lat = 37.775),
                    zoom = 12, scale = 2, size = c(640,640),
                    maptype ='roadmap',
                    color = 'color'))

# Plots the crime points onto the map of Oakland saved to p, I used theme_map
# after finding this: https://rdrr.io/cran/ggthemes/man/theme_map.html

graphic_OK_Google <- p + geom_point(aes(x = lon, y = lat), data = trace_OK, size = 0.5) +
  transition_manual(trace_OK$year) +
  labs(title = "Oakland, CA gun violence Data from 2011-2017",
       subtitle = "Year: {current_frame}",
       caption = "Source: The Trace") +
  ggthemes::theme_map(base_size = 12)

# Saves the animated graph into a gif

anim_save(filename = "trace_OK.gif", animation = graphic_OK_Google, path = "graphics/")
```