---

title: 'Project: Data Science'

author: "Sean E. Curl & Steven Aguilar"

date: "April 25, 2018"

output:

pdf_document: default

html_document: default

word_document: default

---

```{r setup, include=FALSE, warning=FALSE, message=FALSE}

knitr::opts_chunk$set(echo = TRUE)

library(readxl)

library(readr)

library(tidyverse)

library(styler)

library(purrr)

library(ggthemes)

library(ggmap)

library(ggplot2)

library(ggthemes)

library(purrr)

library(ggrepel)

library(revealjs)

```

## About

> - Global Terrorism Database (GTD)

> - Study of Terrorism and Responses to Terrorism (START)

> - University of Maryland

https://www.kaggle.com/START-UMD/gtd

Global Terrorism Database (GTD) is an open-source database with information about terrorist attacks worldwide from 1970 to 2016. The database is maintained by researchers at the National Consortium for the Study of Terrorism and Responses to Terrorism (START), headquartered at the University of Maryland.

#### Definition of terrorism:

> - "The threatened or actual use of illegal force and violence by a non-state actor to attain a political, economic, religious, or social goal through fear, coercion, or intimidation."

#### Observations

> - The data-set included 170,350 observations.

#### Variables?

> - Great than 100 variables on location, tactics, perpetrators, targets, and outcomes.

See the http://start.umd.edu/gtd/downloads/Codebook.pdf for important details on data collection methodology, definitions, and coding schema.

The Global Terrorism Database is funded through START, by the US Department of State and the US Department of Homeland Security Science and Technology Directorate's Office of University Programs. The coding decisions and classifications contained in the database are determined independently by START researchers and should not be interpreted as necessarily representing the official views or policies of the United States Government.

## Load and Clean

```{r echo=FALSE, message=FALSE, warning=FALSE}

terrorism <- read_xlsx(path = "./globalterrorismdb_0617dist.xlsx", col_types = "text") %>%

select("Year" = iyear,

"Month" = imonth,

"Day" = iday,

"Country" = country_txt,

"Region" = region_txt,

"AttackType" = attacktype1_txt,

"Target" = target1,

"Killed" = nkill,

"Wounded" = nwound,

"Summary" = summary,

"Group" = gname,

"Target_Type" = targtype1_txt,

"Weapon_type" = weapsubtype1_txt,

"Motive" = motive,

"City" = city,

"lat" = latitude,

"long" = longitude,

"City" = city) %>%

mutate(Killed = as.numeric(Killed),

Wounded = as.numeric(Wounded), lat = as.numeric(lat), long = as.numeric(long)) %>%

mutate(Casualties = Killed + Wounded) %>%

glimpse()

terrorism %>%

map_dbl(~sum(is.na(.)))

```

In the interested of time and sanity, we selected from the >100 variables, the best variables we thought would lead to a 'quick n' easy' interpretable user experience. As a result, we also created a variable of our own called 'causalities.' The variable is the summation of the Killed and Wounded variables. This variable allowed us a quick and easy way measure the impact a particular terrorist event(s) by using only a single variable.

As you can see, the terrorism database has a bunch of missing observations amongst Target, Killed, Wounded, Summary, Weapon_type, Motive, City, lat, long, and Casualties. Many of these variables are missing simply because the researchers lacked the fidelity for an exact location, weapon, motive, etc. In the case of Killed and Wounded, missing values represent that fact that no person was either killed or wounded in that particular attack. For example, an attack could be deemed a success if the intended target was destroyed (e.g. infrastructure), but there might not have been any casualties as a result of the attack.

## High Impact Areas

```{r echo=FALSE, message=FALSE, warning=FALSE}

Year_highest_attacks <- terrorism %>%

group_by(Year) %>%

summarise(n = n()) %>%

arrange(desc(n)) %>%

slice(1)

Month_highest_attacks <- terrorism %>%

group_by(Month) %>%

summarise(n = mean(n())) %>%

arrange(desc(n)) %>%

slice(1)

Day_highest_attacks <- terrorism %>%

group_by(Day) %>%

summarise(n = mean(n())) %>%

arrange(desc(n)) %>%

slice(1)

country_highest_attacks <- terrorism %>%

group_by(Country) %>%

summarise(n = n()) %>%

arrange(desc(n)) %>%

slice(1)

region_highest_attacks <- terrorism %>%

group_by(Region) %>%

summarise(n = n()) %>%

arrange(desc(n)) %>%

slice(1)

AttackType_highest <- terrorism %>%

group_by(AttackType) %>%

summarise(n = n()) %>%

arrange(desc(n)) %>%

slice(1)

Target_highest <- terrorism %>%

group_by(Target) %>%

summarise(n = n()) %>%

arrange(desc(n)) %>%

slice(1)

highest_killed <- terrorism %>%

group_by(Year, Country, Killed) %>%

summarise(max(Killed)) %>%

arrange(desc(Killed)) %>%

head(1)

highest_wounded <- terrorism %>%

group_by(Year, Country, Wounded) %>%

summarise(max(Wounded)) %>%

arrange(desc(Wounded)) %>%

head(1)

Group_highest <- terrorism %>%

filter(!Group == "Unknown") %>%

group_by(Group) %>%

summarise(n = n()) %>%

arrange(desc(n)) %>%

slice(1)

paste("Year with Highest Terrorist Attacks:", Year_highest_attacks$Year)

paste("Month with Highest Terrorist Attacks:", Month_highest_attacks$Month)

paste("Day with Highest Terrorist Attacks:", Day_highest_attacks$Day)

paste("Country with Highest Terrorist Attacks:", country_highest_attacks$Country)

paste("Region with Highest Terrorist Attacks:", region_highest_attacks$Region)

paste("AttackType with Highest Terrorist Attacks:", AttackType_highest$AttackType)

paste("Target with Highest Terrorist Attacks:", Target_highest$Target)

paste("Maximum peopled killed in an attack are:", "In", highest_killed$Year, highest_killed$`max(Killed)`, "peopled died in",highest_killed$Country)

paste("Maximum peopled wounded in an attack are:", "In", highest_wounded$Year, highest_wounded$`max(Wounded)`, "peopled died in",highest_wounded$Country)

paste("Group with Highest Terrorist Attacks (not unk.):", Group_highest$Group)

```

## Popular Attack Types

```{r echo=FALSE, message=FALSE, warning=FALSE}

attacking_methods <- terrorism %>%

group_by(AttackType) %>%

summarise(n = n()) %>%

arrange(n) %>%

ggplot(mapping = aes(reorder(AttackType, n), n, fill = AttackType)) +

geom_bar(stat = "identity", show.legend = FALSE) +

theme(axis.text.x = element_text(angle = 90, vjust = 0.5, hjust=1, size = 10)) +

ylab("Count") +

xlab("Attack Type") +

ggtitle("Attack Methods Used") +

theme_set(theme_bw()) +

labs(subtitle = "1970 - 2016", caption = "Source: GTD") +

coord_flip()

attacking_methods

```

#### Here's a table view. . .

```{r echo=FALSE, message=FALSE, warning=FALSE}

attacking_methods_table <- terrorism %>%

group_by(AttackType) %>%

summarise(n = n()) %>%

arrange(desc(n))

print(attacking_methods_table)

```

#### Let's take a look at the top three attack types. I'll include Year, Group, and Country too.

```{r echo=FALSE, message=FALSE, warning=FALSE}

attacking_methods_table <- terrorism %>%

filter(!Group == "Unknown") %>%

group_by(AttackType, Year, Country, Group) %>%

summarise(n = max(n())) %>%

arrange(desc(n)) %>%

head(1)

attacking_methods_table2 <- terrorism %>%

filter(AttackType == "Armed Assault") %>%

group_by(AttackType, Year, Country, Group) %>%

summarise(n = max(n())) %>%

arrange(desc(n)) %>%

head(1)

attacking_methods_table3 <- terrorism %>%

filter(AttackType == "Assassination") %>%

group_by(AttackType, Year, Country, Group) %>%

summarise(n = max(n())) %>%

arrange(desc(n)) %>%

head(1)

union1 <- union(attacking_methods_table, attacking_methods_table2)

union2 <- union(union1, attacking_methods_table3)

union2

```

## Number of Terrorist Acivities Year

```{r echo=FALSE, message=FALSE, warning=FALSE}

# Number of Terrorist Acivities Year

count <- terrorism %>%

group_by(Year) %>%

summarise(n = n()) %>%

mutate(Year = as.numeric(Year))

ggplot(count,

aes(x = Year,

y = n, fill = factor(Year))) +

geom_bar(stat = "identity", show.legend = FALSE) +

xlab("Year") +

ylab("Count") +

ggtitle("Global Number of Terrorist Acivities") +

scale_x_continuous(breaks = c(1970, 1980, 1990, 2000, 2010, 2020), limits = c(1970, 2016)) +

theme_set(theme_bw()) +

labs(subtitle = "1970 - 2016",

caption = "Source: GTD") +

coord_flip()

```

#### We can see that the most recent years had a significant increase in violence.

```{r echo=FALSE, message=FALSE, warning=FALSE}

# Number of Terrorist Acivities, Top Six Years

count2 <- terrorism %>%

group_by(Year) %>%

summarise(n = n()) %>%

arrange(desc(n))

print(head(count2))

```

#### Let's take a look at the Year's > 2010.

```{r echo=FALSE, message=FALSE, warning=FALSE}

Year_Great_2010 <- terrorism %>%

filter(!Group == "Unknown") %>%

filter(Year >= '2010' & Year <= '2016') %>%

group_by(Year, Group) %>%

summarise(n = max(n())) %>%

arrange(desc(n))

output1 <- distinct(Year_Great_2010, Country, .keep_all = TRUE)

head(output1, 5)

```

#### How about looking at the Year's < 2010?

```{r echo=FALSE, message=FALSE, warning=FALSE}

Year_Less_2010 <- terrorism %>%

filter(!Group == "Unknown") %>%

filter(Year < '2010') %>%

group_by(Year, Group) %>%

summarise(n = max(n())) %>%

arrange(desc(n))

output <- distinct(Year_Less_2010, Year, .keep_all = TRUE)

head(output, 5)

```

## Terrorist Targets by Count

```{r echo=FALSE, message=FALSE, warning=FALSE}

count <- terrorism %>%

group_by(Target_Type) %>%

summarise(n = n()) %>%

head(10)

ggplot(count,

aes(reorder(x = Target_Type, n),

y = n, fill = factor(Target_Type))) +

geom_bar(stat = "identity", show.legend = FALSE) +

xlab("Target Type") +

ylab("Count") +

ggtitle("Global Terrorist Targets") +

theme(axis.text.x=element_text(angle = 90, hjust =1)) +

theme_set(theme_bw()) +

labs(subtitle = "by Count",

caption = "Source: GTD") +

coord_flip()

```

#### What are the top 5 preferred targets for different groups?

```{r echo=FALSE, message=FALSE, warning=FALSE}

Target_Group <- terrorism %>%

filter(!Group == "Unknown") %>%

group_by(Target_Type, Group) %>%

summarise(n = max(n())) %>%

arrange(desc(n))

output <- distinct(Target_Group, Target_Type, .keep_all = TRUE)

head(output, 5)

```

#### What are top 5 preferred targets within different countries?

```{r echo=FALSE, message=FALSE, warning=FALSE}

Target_Country <- terrorism %>%

group_by(Target_Type, Country) %>%

summarise(n = max(n())) %>%

arrange(desc(n))

output <- distinct(Target_Country, Target_Type, .keep_all = TRUE)

head(output, 5)

```

## Terrorist Attacks by Region

```{r echo=FALSE, message=FALSE, warning=FALSE}

count <- terrorism %>%

group_by(Region) %>%

summarise(n = n())

ggplot(count,

aes(reorder(x = Region, n),

y = n, fill = factor(Region))) +

geom_bar(stat = "identity", show.legend = FALSE) +

xlab("Region") +

ylab("Count") +

ggtitle("Global Terrorist Attacks") +

theme(axis.text.x=element_text(angle = 90, hjust =1)) +

theme_set(theme_bw()) +

labs(subtitle = "by Geographic Region",

caption = "Source: GTD") +

coord_flip()

```

#### It's important to look across a timeframe. Let's compare Region across the years.

```{r echo=FALSE, message=FALSE, warning=FALSE}

Attack.region.count <- terrorism %>%

group_by(Region, Year) %>%

summarise(total = sum(n())) %>%

mutate(Year = as.numeric(Year))

ggplot(Attack.region.count,

aes(x = Year,

y = total, colour = Region, group = Region)) +

geom_line(stat = "identity", lwd = 1) +

xlab("Year") +

ylab("Count") +

ggtitle("Terrorist Attacks") +

scale_x_continuous(breaks = c(1970, 1980, 1990, 2000, 2010, 2020), limits = c(1970, 2016)) +

theme_set(theme_bw()) +

labs(subtitle = "by Region",

caption = "Source: GTD")

```

#### Which regions have the highest amount of terrorist events?

```{r echo=FALSE, message=FALSE, warning=FALSE}

Target_Region <- terrorism %>%

group_by(Year, Region) %>%

summarise(n = max(n())) %>%

arrange(desc(n))

output <- distinct(Target_Region, Year, .keep_all = TRUE)

head(output, 10)

```

#### ...which regions have which groups? Here's the top 6.

```{r echo=FALSE, message=FALSE, warning=FALSE}

Target_Region_Group <- terrorism %>%

filter(!Group == "Unknown") %>%

group_by(Region, Group) %>%

summarise(n = max(n())) %>%

arrange(desc(n))

output <- distinct(Target_Region_Group, Region, .keep_all = TRUE)

head(output, 6)

```

## Terrorist attacks by Group. We'll show the top 10.

```{r echo=FALSE, message=FALSE, warning=FALSE}

count <- terrorism %>%

filter(!Group == "Unknown" & Group %in% c("Taliban", "Shining Path (SL)",

"Islamic State of Iraq and the Levant (ISIL)", "Farabundo Marti National Liberation Front (FMLN)", "Al-Shabaab",

"Irish Republican Army (IRA)",

"Revolutionary Armed Forces of Colombia (FARC)",

"New People's Army (NPA)",

"Kurdistan Workers' Party (PKK)",

"Boko Haram")) %>%

group_by(Group, Year) %>%

summarise(total = n()) %>%

mutate(Year = as.numeric(Year))

ggplot(count,

aes(x = Year,

y = total, colour = Group, group = Group)) +

geom_line(stat = "identity", lwd = 1) +

xlab("Year") +

ylab("Count") +

scale_x_continuous(breaks = c(1970, 1980, 1990, 2000, 2010, 2020), limits = c(1970, 2016)) +

ggtitle("Terrorist Attacks") +

theme(legend.title = element_text(color = "black", size = 10, face = "bold")) +

theme_set(theme_bw()) +

labs(subtitle = "by Group",

caption = "Source: GTD")

```

## Global Maps Plots: Total Terrorism Attacks from 1970 - 2016

```{r echo=FALSE, message=FALSE, warning=FALSE, cache=FALSE}

GTA_graph_data <- terrorism %>%

group_by(Country, long, lat, Casualties) %>%

summarise(total = sum(Casualties))

```

```{r echo=FALSE, message=FALSE, warning=FALSE}

mp <- NULL

mapWorld <- borders("world", colour="gray50", fill="gray50")

mp <- ggplot() + mapWorld +

geom_point(data = GTA_graph_data, mapping = aes(x=long, y=lat,

group = Country, size = total, color = "red"),

show.legend = FALSE) +

theme_bw() +

theme(axis.title = element_blank(),

axis.text = element_blank(),

axis.ticks = element_blank(),

axis.line = element_blank()) +

ggtitle(paste("Global Terror Attacks"))

mp + coord_map(xlim = c(-180, 180)) +

theme(panel.grid.major = element_blank(),

panel.grid.minor = element_blank()) +

labs(subtitle = "1970 - 2016", caption = "Source: GTD")

```

## Global Maps Plots: Terrorism by Groups from 1970 - 2016

```{r echo=FALSE, message=FALSE, warning=FALSE}

test <- terrorism %>%

filter(!Group == "Unknown") %>%

group_by(Group) %>%

summarise(total = n()) %>%

arrange(desc(total)) %>%

slice(1:10)

graph_data1 <- terrorism %>%

filter(!Group == "Unknown" & Group %in% c("Taliban", "Shining Path (SL)",

"Islamic State of Iraq and the Levant (ISIL)", "Farabundo Marti National Liberation Front (FMLN)", "Al-Shabaab",

"Irish Republican Army (IRA)",

"Revolutionary Armed Forces of Colombia (FARC)",

"New People's Army (NPA)",

"Kurdistan Workers' Party (PKK)",

"Boko Haram")) %>%

group_by(Group, long, lat) %>%

summarise(total = n()) %>%

arrange(desc(total))

lp <- NULL

mapWorld <- borders("world", colour="gray50", fill="gray50")

lp <- ggplot(data = graph_data1, mapping = aes(x=long, y=lat)) + mapWorld +

geom_point(aes(color = Group)) +

theme_bw() +

theme(axis.title = element_blank(),

axis.text = element_blank(),

axis.ticks = element_blank(),

axis.line = element_blank()) +

ggtitle("Global Events") +

theme(legend.title = element_text(color = "black", size = 10, face = "bold")) +

guides(color = guide_legend(override.aes = list(size = 1))) +

labs(subtitle = "by the Top 10 Terrorist Organizations since 1970",

caption = "Source: GTD")

lp + coord_map(xlim = c(-180, 180)) +

theme(panel.grid.major = element_blank(),

panel.grid.minor = element_blank())

```

## Terrorist Attacks by City within USA

```{r echo=FALSE, message=FALSE, warning=FALSE}

highest_USA <- terrorism %>%

filter(City %in% c("New York City", "Oklahoma City",

"The Dalles",

"Arlington",

"West",

"Boston",

"Atlanta",

"Orlando")) %>%

group_by(City) %>%

summarise(total = sum(n())) %>%

arrange(desc(total))

ggplot(highest_USA,

aes(reorder(x = City, total),

y = total, fill = factor(City))) +

geom_bar(stat = "identity", width = .5, fill = "red", show.legend = FALSE) +

xlab("City Name") +

ylab("Count") +

ggtitle("Highest Number of U.S. City Terrorist Attacks") +

labs(subtitle = "with New York City",

caption = "Source: GTD") +

theme_set(theme_bw()) +

theme(axis.text.x = element_text(vjust=0.6)) +

coord_flip()

```

#### New York City has the highest number of terrorist attacks since 1970.

```{r echo=FALSE, message=FALSE, warning=FALSE}

count <- terrorism %>%

filter(City == "New York City") %>%

group_by(City, Year) %>%

summarise(total = sum(n())) %>%

arrange(desc(total))

output <- distinct(count, Year, .keep_all = TRUE)

head(output, 10)

```

#### What do these attacks look like across time?

```{r echo=FALSE, message=FALSE, warning=FALSE}

highest_USA <- terrorism %>%

filter(City %in% c("New York City", "Oklahoma City",

"The Dalles",

"Arlington",

"West",

"Boston",

"Atlanta",

"Orlando")) %>%

group_by(City, Year) %>%

summarise(total = sum(n())) %>%

arrange(desc(total)) %>%

mutate(Year = as.numeric(Year))

ggplot(highest_USA,

aes(x = Year,

y = total, colour = City, group = City)) +

geom_line(stat = "identity", lwd = 1) +

xlab("Year") +

ylab("Count") +

scale_x_continuous(breaks = c(1970, 1980, 1990, 2000, 2010, 2020), limits = c(1970, 2016)) +

ggtitle("U.S. City Terrorist Attacks") +

theme(legend.title = element_text(color = "black", size = 10, face = "bold")) +

theme_set(theme_bw()) +

labs(subtitle = "by City",

caption = "Source: GTD")

```

#### Across all U.S. cities, which had the deadliest attacks?

```{r echo=FALSE, message=FALSE, warning=FALSE}

highest_killed_USA <- terrorism %>%

filter(Country == "United States") %>%

group_by(Year, City, Casualties) %>%

summarise(total = sum(Casualties)) %>%

arrange(desc(total))

output <- distinct(highest_killed_USA, City, .keep_all = TRUE)

head(output, 10)

```

#### Let's exclude New York City.

```{r echo=FALSE, message=FALSE, warning=FALSE}

count <- terrorism %>%

filter(City %in% c("Oklahoma City", "Hyder", "Shanksville",

"The Dalles",

"Arlington",

"West",

"Boston",

"Atlanta",

"Orlando")) %>%

group_by(City) %>%

summarise(total = sum(n())) %>%

arrange(desc(total))

ggplot(count,

aes(reorder(x = City, total),

y = total, fill = factor(City))) +

geom_bar(stat = "identity", width = .5, fill = "red", show.legend = FALSE) +

xlab("City Name") +

ylab("Count") +

ggtitle("Terrorist Attacks in Top U.S. Cities") +

theme_set(theme_bw()) +

labs(subtitle = "Minus New York City",

caption = "Source: GTD") +

coord_flip()

```

```{r echo=FALSE, message=FALSE, warning=FALSE}

count <- terrorism %>%

filter(City %in% c("Oklahoma City",

"The Dalles",

"Arlington",

"West",

"Boston",

"Atlanta",

"Orlando")) %>%

group_by(City, Year) %>%

summarise(total = sum(n())) %>%

arrange(desc(total))

output <- distinct(count, City, .keep_all = TRUE)

head(output)

```

#### Again, what do these attacks look like across time?

```{r echo=FALSE, message=FALSE, warning=FALSE}

highest_USA <- terrorism %>%

filter(City %in% c("Oklahoma City",

"The Dalles",

"Arlington",

"West",

"Boston",

"Atlanta",

"Orlando")) %>%

group_by(City, Year) %>%

summarise(total = sum(n())) %>%

arrange(desc(total)) %>%

mutate(Year = as.numeric(Year))

ggplot(highest_USA,

aes(x = Year,

y = total, colour = City, group = City)) +

geom_bar(stat = "identity", lwd = 1) +

xlab("Year") +

ylab("Count") +

scale_x_continuous(breaks = c(1970, 1980, 1990, 2000, 2010, 2020), limits = c(1970, 2016)) +

ggtitle("U.S. City Terrorist Attacks") +

theme(legend.title = element_text(color = "black", size = 10, face = "bold")) +

theme_set(theme_bw()) +

labs(subtitle = "by City (w/o New York City)",

caption = "Source: GTD")

```

#### Across these U.S. cities, which had the deadliest attacks?

```{r echo=FALSE, message=FALSE, warning=FALSE}

highest_killed_USA <- terrorism %>%

filter(Country == "United States") %>%

filter(!City == "New York City") %>%

group_by(Year, City, Casualties) %>%

summarise(total = sum(Casualties)) %>%

arrange(desc(total))

output <- distinct(highest_killed_USA, Year, .keep_all = TRUE)

head(output)

```

#### Shiny App Link:

https://saguilar.shinyapps.io/GTDB/

#### RPubs Presentation:

http://rpubs.com/securl/revealjspresent

#### Presentation created using REVEAL.JS

> - https://revealjs.com/#/

> - https://rmarkdown.rstudio.com/revealjs_presentation_format.html

#### Github Repo:

> - https://github.com/dataSeanC

## QUESTIONS?