1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
|
library(igraph)
setwd('~/Documents/Violence Cascades/Raw Data/')
#================
# (1) load data
#================
#load all three sets of data
arrests <- read.csv("2006to2014arrests2.csv", header=T, colClass=c("character"))
#I need to add the "ir" for this to make sense when I "project"
arrests$ir2 <- paste("ir", arrests$ir_no)
## Match arrests based on date, time, and location
a = arrests[arrests$rd_no=='',]
dtab = table(a$arrest_date)
dates = attr(dtab,'name')[dtab>1]
for (date in dates){
if (which(date==dates)%%10000==0) print(which(date==dates))
ids = which(a$arrest_date==date)
grp = a[ids,]
stab = table(grp$street_nme)
streets = attr(stab,'name')[stab>1]
for (street in streets){
arr_ids = as.numeric(rownames(grp[grp$street_nme==street,]))
arrests$rd_no[arr_ids] = paste('rd',arr_ids[1])
}
}
# now make unique rd_nos for the other people arrested alone
null_rds = which(arrests$rd_no=='')
arrests$rd_no[null_rds] = paste('rd',null_rds)
save(arrests,file='arrests.RData')
#=====================
# (2) Structure Data
#=====================
#get the fields we need for all three: incidents, mni, and "type"
sub.arrests <- subset(arrests, select=c(as.character("rd_no"), as.character('ir2')))
colnames(sub.arrests) <- c("events", "individuals")
#=============================
# (3) Prep for making graphs
#=============================
individuals <- unique(sub.arrests$individuals)
events <- unique(sub.arrests$events)
if (any(individuals %in% events))
stop('vertex name collision')
vertices <- data.frame(name=c(events, individuals),
type=c(rep(FALSE, length(events)),
rep(TRUE, length(individuals))),
stringsAsFactors=FALSE)
#===================================================================
# (4) Make the GRAPH file
#===================================================================
g <- graph.data.frame(sub.arrests, vertices=vertices)
#===================================================================
# Sanity check the resulting igraph object
stopifnot(ecount(g) == nrow(sub.arrests))
stopifnot(vcount(g) == nrow(vertices))
names <- V(g)$name
stopifnot(isTRUE(identical(sort(names), sort(vertices$name))))
inames <- V(g)[V(g)$type]$name
stopifnot(isTRUE(identical(sort(inames), sort(individuals))))
enames <- V(g)[! V(g)$type]$name
stopifnot(isTRUE(identical(sort(enames), sort(events))))
#===================================================================
#now do the converstion into a single network
net1 <- bipartite.projection(g)
person <- net1[[2]]
#===================================================================
# (5) Define attributes on nodes
#===================================================================
# set attributes from arrests file
attribs <- arrests
match_vector = match(V(person)$name, attribs$ir2)
V(person)$sex <- as.character(attribs$sex_code_cd[match_vector])
V(person)$race <- as.character(attribs$race_code_cd[match_vector])
# V(person)$age <- as.character(attribs$age[match_vector])
V(person)$dob <- as.character(as.Date(attribs$birth_date[match_vector],format='%m/%d/%Y'))
# January 1, 2006 is Day 1 of the study period
start_date = as.Date("2005-12-31")
## Get first arrest date in the study period for each person
sub.arrests$dates = as.Date(arrests$arrest_date,format='%m/%d/%Y')
sub.arrests = sub.arrests[order(sub.arrests$dates),]
#===================================================================
# get victim attributes
shootings <- read.csv("shooting-data-withdate2.csv", header = T)
victims = shootings[shootings$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)
# set victim data in network
vtab = as.data.frame(table(victims$ir2))
match_vector = match(V(person)$name,vtab$Var1)
V(person)$vic.nonfatal = vtab$Freq[match_vector]
V(person)$vic.nonfatal[is.na(V(person)$vic.nonfatal)] = 0
V(person)$vic.fatal = V(person)$name %in% murders$ir2
V(person)$vic = V(person)$name %in% union(victims$ir2,murders$ir2)
# add fatal shooting dates to the network
match_vector = match(V(person)$name, murders$ir2)
fatal_dates = murders$INJURY_DATE[match_vector]
fatal_dates = as.character(as.Date(fatal_dates,format='%m/%d/%y'))
V(person)$fatal_date = fatal_dates
# add nonfatal shooting dates to the network
match_vector = match(victims$ir2,V(person)$name)
vics = which(V(person)$vic.nonfatal>0)
nfd1 = nfd2 = nfd3 = nfd4 = nfd5 = rep(0,length(vics))
for(i in 1:length(vics)){
if (i%%3000==0) print(i)
name = vics[i]
ids = which(match_vector==name)
dates = sort(as.Date(victims$INCIDENT_DATE[ids],format='%m/%d/%y'))
nfd1[i] = as.character(dates[1])
nfd2[i] = as.character(dates[2])
nfd3[i] = as.character(dates[3])
nfd4[i] = as.character(dates[4])
nfd5[i] = as.character(dates[5])
}
V(person)$nonfatal_date_1 = NA
V(person)$nonfatal_date_2 = NA
V(person)$nonfatal_date_3 = NA
V(person)$nonfatal_date_4 = NA
V(person)$nonfatal_date_5 = NA
V(person)$nonfatal_date_1[vics] = nfd1
V(person)$nonfatal_date_2[vics] = nfd2
V(person)$nonfatal_date_3[vics] = nfd3
V(person)$nonfatal_date_4[vics] = nfd4
V(person)$nonfatal_date_5[vics] = nfd5
# convert dates into numeric values ("days")
start_date
V(person)$fatal_day = as.numeric(as.Date(V(person)$fatal_date)-start_date)
V(person)$nonfatal_day_1 = as.numeric(as.Date(V(person)$nonfatal_date_1)-start_date)
V(person)$nonfatal_day_2 = as.numeric(as.Date(V(person)$nonfatal_date_2)-start_date)
V(person)$nonfatal_day_3 = as.numeric(as.Date(V(person)$nonfatal_date_3)-start_date)
V(person)$nonfatal_day_4 = as.numeric(as.Date(V(person)$nonfatal_date_4)-start_date)
V(person)$nonfatal_day_5 = as.numeric(as.Date(V(person)$nonfatal_date_5)-start_date)
#===================================================================
# set gang attributes
gangs <- read.csv("Sept2014-ganglist.csv", header=T)
gangs = gangs[match(unique(gangs$IR_NO),gangs$IR_NO),]
gangs$ir2 <- paste("ir", gangs$IR_NO)
V(person)$gang.member <- V(person)$name %in% gangs$ir2
match_vector = match(V(person)$name, gangs$ir2)
gnames = gangs$GANG_NAME[match_vector]
gnames = as.character(gnames)
gnames[V(person)$gang.member==''] = 'Unknown'
gnames[V(person)$gang.member==F] = 'None'
V(person)$gang.name <- as.character(gnames)
V(person)$faction.name <- as.character(gangs$FACTION_NAME[match_vector])
#===================================================================
# create id number
# save data
# person = remove.edge.attribute(person,'weight')
# person_data = get.data.frame(person,'both')
save(person, file="chi-19aug2015.RData")
#===================================================================
# get LCC of the network
lcc = induced.subgraph(person,which(clusters(person)$membership==which.max(clusters(person)$csize)))
V(lcc)$id = rank(V(lcc)$name)
V(lcc)$ir_no = V(lcc)$name
V(lcc)$name = V(lcc)$id
vic_ids = which(V(lcc)$vic)
lcc_edges = as_data_frame(lcc,'edges')
# update lcc_verts
lcc_verts = get.data.frame(lcc,'vertices')
lcc_verts = lcc_verts[,c(1,23,24,2:22)]
# save file
save(lcc, lcc_verts, lcc_edges, vic_ids, file="lcc.RData")
#####
# old stuff
lcc_data = get.data.frame(lcc,'both')
lcc = set.vertex.attribute(graph=lcc, name='name', value=V(lcc)$id)
row.names(lcc_data$vertices) = lcc_data$vertices$id
vertices = lcc_data$vertices[c('name','vic','vic.fatal','vic.nonfatal',
'fatal_day','nonfatal_day_1','nonfatal_day_2',
'nonfatal_day_3','nonfatal_day_4','nonfatal_day_5')]
write.csv(vertices,file='lcc_vertices.csv')
write.csv(lcc_data$edges,file='lcc_edges.csv')
lcc = graph.data.frame(lcc_edges, directed=FALSE,
vertices=lcc_verts[,c('name','vic','vic.fatal','vic.nonfatal')])
write.graph(lcc,'lcc.gml','gml')
|
