Intro

In this investigation, I use the 5 years worth of freely available data from massshootingtracker.org (all the observations of which have news sources), and visualize 5 years of incidents. I work with:

To be completed:

Data

An initial look at the data, we have (5 years worth) from

#initial
dim(dfull)
## [1] 1890   13
names(dfull)
##  [1] "date"                        "name_semicolon_delimited"   
##  [3] "killed"                      "wounded"                    
##  [5] "city"                        "state"                      
##  [7] "sources_semicolon_delimited" "week"                       
##  [9] "year"                        "week2"                      
## [11] "month"                       "month2"                     
## [13] "dateF"

The data contains 1890 observations (shootings) and 12 variables for each shooting.

#initial1
dfull %>% arrange(desc(dateF)) %>% select(city, date, killed, wounded) %>% head()
##                 city      date killed wounded
## 1 Sutherland Springs 11/5/2017     27      20
## 2            Detroit 11/5/2017      0       4
## 3         Youngstown 11/4/2017      0       4
## 4             Austin 11/4/2017      0       4
## 5       Santa Monica 11/4/2017      1       4
## 6            Gardena 11/4/2017      0       4

The most recent 6 shootings in the data I downloaded are shown above.

Initial Look

Number of incidents aggregated by year

Aggregated by month

Visualizing just the sheer number of incidents aggregated by month we notice (i) a certain amount of inherent variability and (ii) a distinctive uptrend, NB we exclude November/December 2017 due to lack of full results

The number wounded and killed each month