Overview

The goal of this work is to get acquinented with the Leaflet framework which allows to create and customize interactive maps using the ‘Leaflet’ JavaScript library and the ‘htmlwidgets’ package. The maps can be used directly from the R console, from ‘RStudio’, in Shiny applications and R Markdown documents.

More information can be retrieved from: https://cran.r-project.org/web/packages/leaflet/index.html

In this work we will find cities with the highest crime rates for every state from data set and and show them on the USA map which will contain marks and circles elements with radiuses depended on the value of the crime rate.

For the work we will use data from two sources:

Loading the files with the data

library(dplyr)

offenceReportUrl <- "https://ucr.fbi.gov/crime-in-the-u.s/2018/preliminary-report/tables/table-4/table-4.xls"
geoPositionsUrl <- "https://www.downloadexcelfiles.com/sites/default/files/docs/lat-long-cities-usa-1507j.xlsx"

for(url in c(offenceReportUrl, geoPositionsUrl)) {
  fileName <- URLdecode(url) %>% basename()
  if (!file.exists(fileName)) {
    download.file(url, fileName, method = "curl")
    cat("\n=====>The file [", fileName, "] was downloaded successfully")
  } else {
    cat("\n=====>The file [", fileName, "] already exists!")
  }
}

Read the files into dataframes

#install.packages("readxl")
library("readxl")

offenceReportFileName <- URLdecode(offenceReportUrl) %>% basename()
geoPositionsFileName <- URLdecode(geoPositionsUrl) %>% basename()

# read files
offenceReportDataFrame <- read_excel(offenceReportFileName, skip=4)
geoPositionsDataFrame <- read_excel(geoPositionsFileName)

Data preparation

colnames(offenceReportDataFrame)[3] <- "Year"
colnames(offenceReportDataFrame)[4] <- "Population"
colnames(offenceReportDataFrame)[5] <- "Violent_Crime"


#install.packages("dplyr")
library(dplyr)

offenceReportDataFrame <- select(filter(offenceReportDataFrame, Year == 2017), c(State, City, Year, Population, Violent_Crime))

stateList <- offenceReportDataFrame$State

notEmptyName <- "foobar"
for(i in 1 : length(stateList)) {
  if(!is.na(stateList[i])) {
    notEmptyName <- stateList[i]
  } else {
    stateList[i] <- notEmptyName
  }
}

offenceReportDataFrame$State <- stateList

tidyOffenceReportDataFrame <- offenceReportDataFrame

Find cities with the highiest crime rate

# Calculate crime rate for every city
tidyOffenceReportDataFrame$Crime_Rate <- tidyOffenceReportDataFrame$Violent_Crime / tidyOffenceReportDataFrame$Population

# find the most crimiest city in every state
tidyOffenceReportDataFrame <- merge(aggregate(Crime_Rate ~ State, tidyOffenceReportDataFrame, max), tidyOffenceReportDataFrame, by = c("Crime_Rate"))

tidyOffenceReportDataFrame <- subset(tidyOffenceReportDataFrame, select = -c(3)) 
tidyOffenceReportDataFrame <- tidyOffenceReportDataFrame[,c(2, 3, 4, 1, 5)]
colnames(tidyOffenceReportDataFrame)[1] <- "State"

# sort by crime rate
tidyOffenceReportDataFrame <- tidyOffenceReportDataFrame[order(-tidyOffenceReportDataFrame$Crime_Rate), ]


# because of presence of reference numbers in the report, we need to get rid of them
tidyOffenceReportDataFrame$City <- replace(as.character(tidyOffenceReportDataFrame$City), tidyOffenceReportDataFrame$City == "NEW ORLEANS7", "NEW ORLEANS")
tidyOffenceReportDataFrame$City <- replace(as.character(tidyOffenceReportDataFrame$City), tidyOffenceReportDataFrame$City == "HONOLULU5", "HONOLULU")
tidyOffenceReportDataFrame$City <- replace(as.character(tidyOffenceReportDataFrame$City), tidyOffenceReportDataFrame$City == "HOUSTON5", "HOUSTON")
tidyOffenceReportDataFrame$City <- replace(as.character(tidyOffenceReportDataFrame$City), tidyOffenceReportDataFrame$City == "NASHVILLE METROPOLITAN", "NASHVILLE")

crimeRatesDataFrame <- tidyOffenceReportDataFrame

Match names of the states and cities in the dataset with the names of them in the geo data set

geoPositionsDataFrame$State <- toupper(geoPositionsDataFrame$State)
geoPositionsDataFrame$City <- toupper(geoPositionsDataFrame$City)

# merge two data sets
crimeRatesDataFrame <- merge(crimeRatesDataFrame, geoPositionsDataFrame, by = c("State", "City"))

# split geo coordinates
splitGeo <- data.frame(do.call('rbind', strsplit(as.character(crimeRatesDataFrame$Location),' ',fixed=TRUE)))
colnames(splitGeo) <- c('Latitude', 'Longitude')

# extract geo coordinates
#install.packages("stringr")
library("stringr")

splitGeo$Latitude <- as.numeric(str_extract(splitGeo$Latitude, "\\d+\\.*\\d*"))
splitGeo$Longitude <- 0 - as.numeric(str_extract(splitGeo$Longitude, "\\d+\\.*\\d*"))

crimeRatesDataFrame <- cbind(crimeRatesDataFrame, splitGeo)

geoCrimeRatesDataFrame <- crimeRatesDataFrame

Show data on a map

#install.packages('leaflet')
library(leaflet)

map <- leaflet() %>% addTiles()
map <- map %>% 
    addMarkers(lat = geoCrimeRatesDataFrame$Latitude, 
               lng = geoCrimeRatesDataFrame$Longitude, 
               popup = paste(geoCrimeRatesDataFrame$City,
                             "<br>State: ", formatC(geoCrimeRatesDataFrame$State),
                             "<br>Population, people: ", formatC(geoCrimeRatesDataFrame$Population, format = "d", big.mark = ","),
                             "<br>Crime rate, per 100k people: ", formatC(geoCrimeRatesDataFrame$Crime_Rate * 100000, format = "d", big.mark = ",")
                             )
               ) %>%
    addCircles(lat = geoCrimeRatesDataFrame$Latitude,
               lng = geoCrimeRatesDataFrame$Longitude,
               weight = 1, 
               radius = geoCrimeRatesDataFrame$Crime_Rate * 2 * 10^7,
              )
map

Cities with the highest crime rate per state