###################
# load crime data
library(xlsx)
setwd("~/Documents/Violence Cascades/Raw Data/")
data = read.xlsx('Chicago-crime-data.xlsx',1)

# crime rate
plot(data$Year,data$Index.Rate.per.100.000,type='l',col='#1f78b4',lwd=2,
     xlab='',ylab='Rate per 100,000',cex.lab=.6,cex.axis=.6)

# homocide rate
plot(data$Year,data$Homicide.Rate,type='l',col='#1f78b4',lwd=2,
     xlab='',ylab='Rate per 100,000',cex.lab=.6,cex.axis=.6)

###################
# load shootings data
# for lcc
load('lcc.RData')
vic_dates = as.Date(unlist(lcc_verts[,10:15]))
vic_dates = vic_dates[!is.na(vic_dates)]
vdh = hist(vic_dates, breaks='months')

plot(vdh$mids,vdh$counts,type='l',col='#1f78b4',lwd=2,
     xlab='',ylab='Shootings',xaxt='n',cex.lab=.6,cex.axis=.6)
axis(1,at=vdh$breaks[seq(1,102,12)],
     lab=2006:2014,cex.axis=.6)


# for all recorded shootings
shootings <- read.csv("shooting-data-withdate2.csv", header = T)
victims = shootings[shootings$INV_PARTY_TYPE_CD=="VIC",]
victims = victims[!is.na(victims$IR_NO),]
victims$ir2 <- paste("ir", victims$IR_NO)

# get murder victim attributes
murders = read.csv("murder-victims-13nov.csv", header=T)
murders = murders[!is.na(murders$VICTIM_IR_NO),]
murders = murders[murders$INJURY_DESCR=="SHOT",]
murders = murders[match(unique(murders$VICTIM_IR_NO),murders$VICTIM_IR_NO),]
murders = murders[as.Date(murders$INJURY_DATE,format='%m/%d/%y')>=start_date,]
murders$ir2 = paste("ir", murders$VICTIM_IR_NO)

# clear nonfatals that led to death
v = victims[victims$IR_NO %in% murders$VICTIM_IR_NO,]
rows = c()
for(i in 1:dim(v)[1]){
  row = which(rownames(victims)==as.numeric(rownames(v[i,])))
  m = murders[murders$VICTIM_IR_NO==v$IR_NO[i],]
  dup = as.Date(v$INCIDENT_DATE[i],format='%m/%d/%y') %in% as.Date(m$INJURY_DATE,format='%m/%d/%y')
  if(dup==T) rows = c(rows,row)
}
victims = victims[-rows,]

vic_dates = c(as.Date(murders$INJURY_DATE,format='%m/%d/%y'),as.Date(victims$INCIDENT_DATE,format='%m/%d/%y'))