---
title: "ANLY 512 Final - US MASS SHOOTING"
author: "Yiran Huang"
date: "4/21/2020"
output:
  flexdashboard::flex_dashboard:
    storyboard: yes
    source: embed
  html_document:
    df_print: paged
---

```{r setup, include=FALSE}
library(data.table)
library(readr)
library(plotly)
library(ggplot2)
library(maps)
library(tm)
library(wordcloud)
library(IRdisplay)
library(knitr)
library(shiny)
```

Project Introduction
===================================== 

Working as a researcher for justice at John Jay College, I always do lots of researchers on criminal incidents. In the decade between 2010 and 2020, the horrific scenes of mass shootings have haunted the nation’s collective conscience. Each breaking news alert floods the nation with grief and anger at this senseless, preventable violence. This statistic shows the number of mass shootings in the United States between 1966 and February 26, 2020,

*Definition of a mass shooting from wikipedia:*
A mass shooting is an incident involving multiple victims of gun violence. There is no widely accepted definition of the term mass shooting. The United States FBI defines a "mass murder" as "four or more murdered during an event with no "cooling-off period" between the murders." Based on this, it is generally agreed that a mass shooting is whenever four or more people are shot (injured or killed), not including the shooter(s).
Different media outlets and research groups use different definitions for the term "mass shooting" For example, crime violence research group Gun Violence Archive defines a "mass shooting" as "four or more shot (injured or killed) in a single incident, at the same general time and location, not including the shooter,” differentiating between mass shooting and mass murder and not counting shooters as victims.
The United States’ Congressional Research Service acknowledges that there is not a broadly accepted definition and defines a "public mass shooting"as an event where someone selects four or more people and shoots them with firearms in an indiscriminate manner, echoing the FBI's definition of the term "mass murder", but adding the indiscriminate factor.

This analysis reflects a compilation of 55 years (1966 to 2020) of original data on mass shootings in the United States, sourced from my work data. These records enable unique insights into the circumstances of mass shootings. For full details on data collection, see our methodology.

***Number of columns: 50***
***Number of rows: 348***

Data manipulation - Since this is sensitive work data, my employer asked me to anonymize some of the data.When I’m replicating this for work, all types of data are included and the actual industry values are retained.Mass shootings are not a random, inevitable element of American life today. Rather, this report illuminates trends wish can help point lawmakers to strategies to curb these tragedies.

Therefore, I would like to looking into 6 aspects of US mass shooting data. I’ll be making data visulized and trying to find out the answer to the following 6 questions:

**Question 1: Have mass shootings in US been increasing over the years?**

Mass shootings have been resulting in more people shot in each incident in recent years.

The graph illustrates that Mass shootings have not only increased in frequency in recent years, but also the death and injury counts of each shooting have been rising.In the decade between 2012 and 2019, there are more than 10 shootings happened each year. These numbers are staggering, yet they represent just a small portion of the lives forever changed after a mass shooting shakes a community with terror and grief.

In nearly all mass shootings over this period, the shooter was an adult man who acted alone.


Victims Trends in the Mass Shooting Cases in the US

**Question 2: Have victims in mass shooting cases increasing over some years?**

**Question 3:Are some states more prone to shootings?**

**Question 4: Is there a relationship between mental illness and the mass shooting cases?**

**Question 5:  What was the distribution characteristics of victims by race?**

**Question 6:What was the distribution characteristics of shooters by age and race?**



```{r}
setwd("C:/Users/Richie/Desktop/R hw/hw2/")
MS_dataset <- read_csv("C:/Users/Richie/Desktop/R hw/hw2/Mass Shootings Dataset.csv"
                              , col_types = cols(Date = col_date(format = "%m/%d/%Y")))
MS_dataset <- data.table(MS_dataset)

MS_dataset[,Month:=as.factor(month(Date))]
MS_dataset[,Year_n:=as.numeric(year(Date))]
MS_dataset[,Year:=as.factor(year(Date))]

MS_dataset[Gender=='M',Gender:="Male"]
MS_dataset[Gender=='M/F',Gender:="Male/Female"]
MS_dataset[is.na(Gender),Gender:="Unknown"]
MS_dataset[,Gender:=as.factor(Gender)]
```


Mass Shooting Trends in America
===================================== 

Row {data-width=1000}
-------------------------------------

### Figure1: Mass Shootings in US by years and month {data-commentary-width=400}

```{r, echo=FALSE}
plot_ly(data = MS_dataset
        ,type = 'scatter'
        ,mode = 'markers' 
        ,hoverinfo = 'text'
        ,x = ~Month
        ,y = ~Year
        ,size = ~`Total victims`
        ,color = ~Gender
        ,colors = c('Red', 'Blue', 'Green', 'Black')
        ,alpha = 0.6
        ,text = ~paste("Location: ", Location
               ,'\n Date: ', Date 
               ,'\n Total victims : ', `Total victims` 
               ,'\n Fatalities : ', Fatalities
               ,'\n Injured : ', Injured)) %>% 
  layout(title = "Mass Shootings in US by years and month"
         , xaxis = list(title = "Month")
         , yaxis = list(title = "Years"))
```

Row {data-width=1000}
-------------------------------------

### Figure 2 Number of incidents by years

```{r}
plot_ly(data = MS_dataset
        ,type = 'histogram'
        ,mode = 'markers'
        ,x = ~Year
        ,alpha = 0.9) %>% 
  layout(title = "Number of incidents by years"
         , xaxis = list(title = "")
         , yaxis = list(title = "Number of incidents"))
```

### Analysis

Mass shootings have been resulting in more people shot in each incident in recent years.

The graph illustrates that Mass shootings have not only increased in frequency in recent years, but also the death and injury counts of each shooting have been rising.In the decade between 2012 and 2019, there are more than 10 shootings happened each year. These numbers are staggering, yet they represent just a small portion of the lives forever changed after a mass shooting shakes a community with terror and grief.

In nearly all mass shootings over this period, the shooter was an adult man who acted alone.




Victims Trends in the Mass Shooting Cases in the US
===================================== 

Row {data-width=1000}
-------------------------------------

```{r}
# Font Settings 
f1 <- list(
  family = "Arial, sans-serif",
  size = 14,
  color = "grey"
)
f2 <- list(
  family = "Old Standard TT, serif",
  size = 12,
  color = "black"
)

# Axis settings
ax <- list(
  title = "Month",
  titlefont = f1,
  showticklabels = TRUE,
  tickangle = 0,
  tickfont = f2,
  exponentformat = "E"
)

ay <- list(
  title = "Year",
  titlefont = f1,
  showticklabels = TRUE,
  tickangle = 0,
  tickfont = f2,
  exponentformat = "E"
)

b1 <- list(
  text = "Total victims",
  font = f1,
  xref = "paper",
  yref = "paper",
  yanchor = "bottom",
  xanchor = "center",
  align = "center",
  x = 0.5,
  y = 1,
  showarrow = FALSE
)

b2 <- list(
  text = "Injured",
  font = f1,
  xref = "paper",
  yref = "paper",
  yanchor = "bottom",
  xanchor = "center",
  align = "center",
  x = 0.5,
  y = 1,
  showarrow = FALSE
)

b3 <- list(
  text = "Fatalities",
  font = f1,
  xref = "paper",
  yref = "paper",
  yanchor = "bottom",
  xanchor = "center",
  align = "center",
  x = 0.5,
  y = 1,
  showarrow = FALSE
)


hm1 <- 
#plot_ly(data = MS_dataset[`S#`!=1,]
plot_ly(data = MS_dataset
        ,type = 'heatmap'
        ,colors = colorRamp(c("yellow", "blue", "darkred"))
        ,x = ~Month
        ,y = ~Year
        ,z = ~`Total victims`)%>%
  layout(showlegend = T
         , xaxis = ax
         , yaxis = ay
         , annotations = b1)

hm2 <- 
#plot_ly(data = MS_dataset[`S#`!=1,]
plot_ly(data = MS_dataset
        ,type = 'heatmap'
        ,colors = colorRamp(c("grey", "darkgrey", "black"))
        ,x = ~Month
        ,y = ~Year
        ,z = ~`Injured`)%>%
  layout(showlegend = T
         , xaxis = ax
         , annotations = b2
         , yaxis = list(
  title = "",
  zeroline = FALSE,
  showline = FALSE,
  showticklabels = FALSE,
  showgrid = FALSE
))

hm3 <- 
#plot_ly(data = MS_dataset[`S#`!=1,]
plot_ly(data = MS_dataset
        ,type = 'heatmap'
        ,colors = colorRamp(c("orange", "darkred", "black"))
        ,x = ~Month
        ,y = ~Year
        ,z = ~Fatalities)%>%
  layout(showlegend = T
         , xaxis = ax 
         , annotations = b3
         , yaxis = list(
  title = "",
  zeroline = FALSE,
  showline = FALSE,
  showticklabels = FALSE,
  showgrid = FALSE
))

fig3 <- subplot(hm1, hm2, hm3)
fig3
```



Row {data-width=1000}
-------------------------------------

```{r}
fig4  <- plot_ly(data = MS_dataset
        ,type = 'bar'
        ,mode = 'markers' 
        ,hoverinfo = 'text'
        ,x = ~Year
        ,y = ~ `Total victims` 
        ,color = 'Red'
        ,alpha = 0.9
        ,text = ~paste(
               'Fatalities : ', Fatalities
               ,'\n Injured : ', Injured
)) %>% 
  layout(title = "Number of Total victims by years"
         , xaxis = list(title = "")
         , yaxis = list(title = "Number of victims"))

fig4

```

### Analysis
The number of incidents, fatalities and injuries have only increased over the years with an alarming spike after 2014. Back in 2015, journalists at the New York Times reported that on average, shootings that left more than 4 or more people wounded or dead occurred more than once a day. This is a terrifying statistic to think about.



Geography of Mass Shooting in US
===================================== 

Row {data-width=1000}
-------------------------------------

```{r}
g <- list(
    scope = 'usa'
  , projection = list(type = 'albers usa')
  , showland = TRUE
  , landcolor = 'grey'
  , subunitwidth = 1
  , countrywidth = 1
#  , subunitcolor = toRGB("white")
#  , countrycolor = toRGB("white")
)

plot_geo(MS_dataset
         #, locationmode = 'USA-states'
         , sizes = c(10, 300)) %>%
  add_markers(
    x = ~Longitude
    , y = ~Latitude
    , size = ~`Total victims`
    , color = ~Fatalities
    , colors = colorRamp(c("yellow", "red", "black"))
    , hoverinfo = "text"
    , text = ~paste(MS_dataset$Title
                    , '\n Fatalities: ', MS_dataset$Fatalities
                    , '\n Injured: ', MS_dataset$Injured)
  ) %>%
  layout(title = 'Geography of Mass Shooting in US', geo = g)
```

Row {data-width=1000}
-------------------------------------


```{r}
MS_dataset$State <- sapply(MS_dataset$Location, function(x){
  temp <- strsplit(x, split = ",")
  sapply(temp, function(y){y[2]
    
  })
})

colors_pie1 <- c('rgb(211,94,96)', 'rgb(128,133,133)', 'rgb(144,103,167)', 'rgb(171,104,87)', 'rgb(114,147,203)')
plot_ly(data = MS_dataset[!is.na(State),.('Number of incidents'= uniqueN(`S#`)),by=State]
        ,type = 'pie'
        ,labels = ~State
        ,values = ~`Number of incidents`
        ,textposition = 'inside'
        ,insidetextfont = list(color = '#FFFFFF')
        ,marker = list(colors = colors_pie1,
                      line = list(color = '#FFFFFF', width = 1)))%>%
  layout(title = "Number of incidents by States",
         showlegend = T)
```

### Analysis
United States is experiencing a gun epidemic across the country. With incidents spanning over the past 6 decades, we can see the state-wise distribution of incidents, fatalities and total victim count. we see that California has a high number of incidents and victims overall, the second is Florida. 


Mental Health Issues
===================================== 


```{r}
MS_dataset[`Mental Health Issues`=="unknown",`Mental Health Issues`:="Unknown"]


plot_ly(data = MS_dataset[,.(`Total victims`,`Mental Health Issues`)]
        ,type = 'pie'
        ,labels = ~`Mental Health Issues`
        ,values = ~`Total victims`
        ,textposition = 'inside'
        ,insidetextfont = list(color = '#FFFFFF')
        ,marker = list(colors = colors_pie1,
                      line = list(color = '#FFFFFF', width = 1)))%>%
  layout(title = "Mental Health Issues",
         showlegend = T)
```

*** 
## Analysis
Four assumptions frequently arise in the aftermath of mass shootings in the United States: (1) that mental illness causes gun violence, (2) that psychiatric diagnosis can predict gun crime, (3) that shootings represent the deranged acts of mentally ill loners, and (4) that gun control “won’t prevent” another Newtown (Connecticut school mass shooting). Each of these statements is certainly true in particular instances. Yet, as we show, notions of mental illness that emerge in relation to mass shootings frequently reflect larger cultural stereotypes and anxieties about matters such as race/ethnicity, social class, and politics. These issues become obscured when mass shootings come to stand in for all gun crime, and when “mentally ill” ceases to be a medical designation and becomes a sign of violent threat.



Total Victims by Race
===================================== 


```{r}
plot_ly(data = MS_dataset[,.('Total victims'= sum(`Total victims`)),by=.(Race,Year)]
        ,type = 'bar'
        ,mode = 'markers'
        ,x = ~Year
        ,y = ~`Total victims`
        ,color =~Race
        ,alpha = 0.9) %>% 
  layout(title = "Total victims by Race"
         , showlegend = T
         , barmode = 'stack'
         , position = 1
         , xaxis = list(title = "")
         , yaxis = list(title = "")
         , legend = list(x = 0, y = 1)
         , hovermode = 'compare')
```

***
## Analysis
The high number of tragic mass shootings that have occurred in the United States has led to a large amount of attention on the profile of the people who commit such violent acts. A look at the worst mass shootings in the United States suggests no clear common connections of races in a tendency to undertake mass shootings. 


Shooter's Information Analysis
===================================== 


```{r}
a1 <- 
ggplot(data = MS_dataset[!is.na(Age)&Age!=0&Age<=70,], aes(x = Race, y = Age)) +
  geom_boxplot(aes(col = Race)) + 
  ggtitle("Age of the shooter & Race") +
  labs(x = "Race", y = "Age") +

  theme(axis.text.x = element_text(angle = 0
                                   , size = 9
                                   , color = 'black'
                                   , hjust = 1),
        legend.position="none") +
  geom_hline(aes(yintercept = median(Age))
             , colour = 'red'
             , linetype = 2
             , alpha = 0.5) + 
  geom_hline(aes(yintercept = mean(Age))
             , colour = 'blue'
             , linetype = 2
             , alpha = 0.5)

ggplotly(
a1
)
```

*** 
## Analysis 
The race of the perpetrators in mass shootings is cause for widespread media coverage. Throughout the last six decades, we observed racial diversity in the perpetrators, however, through our visual exploration of the data set it is clear that White perpetrators are predominant among all other races.