forked from topepo/useR2016
-
Notifications
You must be signed in to change notification settings - Fork 0
/
okc_data.R
175 lines (144 loc) · 6.73 KB
/
okc_data.R
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
###################################################################
## Code for the useR 2016 tutorial "Never Tell Me the Odds! Machine
## Learning with Class Imbalances" by Max Kuhn
##
## Slides and this code can be found at
## https://github.com/topepo/useR2016
##
## packages used here are: lubridate
##
## Data are at: https://github.com/rudeboybert/JSE_OkCupid
library(lubridate)
###################################################################
## Some levels of predictors have spaces or symbols; fix these
## and replace "" with "missing"
fix_levels <- function(x) {
x <- gsub("’", "", x)
x <- gsub("[[:space:]]", "_", x)
x <- gsub("[[:punct:]]", "_", x)
x <- gsub("__", "_", x, perl = TRUE)
x <- gsub("__", "_", x, perl = TRUE)
x <- gsub("__", "_", x, perl = TRUE)
x[x == ""] <- "missing"
resort_lvl(x)
}
###################################################################
## resort categorical predictors so that "missing" is the first level
resort_lvl <- function(x) {
x <- as.character(x)
lv <- c("missing", sort(unique(x[x != "missing"])))
factor(x, levels = lv)
}
###################################################################
## Unpackage the data and read in. A compressed version of the csv
## file is at https://github.com/rudeboybert/JSE_OkCupid
raw <- read.csv("profiles.csv",
stringsAsFactors = FALSE)
raw <- raw[, !grepl("^essay", names(raw))]
###################################################################
## Compute the number of days since last online
tmp_last <- ymd(substring(raw$last_online, 1, 10))
tmp_last <- difftime(max(tmp_last), tmp_last, units = "days")
raw$last_online <- as.numeric(tmp_last)
###################################################################
## encode the "easy" categorical predictors
raw$body_type <- fix_levels(raw$body_type)
raw$drinks <- fix_levels(raw$drinks)
raw$drugs <- fix_levels(raw$drugs)
raw$education <- fix_levels(raw$education)
raw$diet <- fix_levels(raw$diet)
raw$job <- fix_levels(raw$job)
raw$offspring <- fix_levels(raw$offspring)
raw$pets <- fix_levels(raw$pets)
raw$orientation <- fix_levels(raw$orientation)
raw$sex <- fix_levels(raw$sex)
raw$smokes <- fix_levels(raw$smokes)
raw$drugs <- fix_levels(raw$drugs)
raw$status <- fix_levels(raw$status)
###################################################################
## Income is basically encoded categorical so we will make it a factor
test <- ifelse(raw$income == -1, NA, raw$income)
test <- factor(paste0("inc", test), levels = c("missing", paste0("inc", sort(unique(test)))))
test[is.na(test)] <- "missing"
raw$income <- test
###################################################################
## Split their location into city and state. There are some R functions
## (ahem, randomForest) that can only handle predictors with <=52
## levels so we take the long tail of the distribution and truncate
## some cities to "other"
tmp_where <- strsplit(raw$location, split = ", ")
where_state <- unlist(lapply(tmp_where, function(x) if(length(x) == 2) x[2] else "missing"))
where_town <- unlist(lapply(tmp_where, function(x) if(length(x) == 2) x[1] else "missing"))
town_tab <- sort(-table(where_town))
where_town[!(where_town %in% names(town_tab)[1:50])] <- "other"
raw$where_state <- factor(gsub(" ", "_", where_state))
raw$where_town <- factor(gsub(" ", "_", where_town))
raw$location <- NULL
###################################################################
## Some predictors have values and modifiers that describe how
## serious they are about their choice. We will create predictors
## for both characteristics of their answer
## for religon, split religon and modifier
tmp_relig_split <- strsplit(raw$religion, split = " ")
tmp_relig <- unlist(lapply(tmp_relig_split, function(x) x[1]))
tmp_relig[tmp_relig == ""] <- "missing"
tmp_relig[is.na(tmp_relig)] <- "missing"
raw$religion <- resort_lvl(tmp_relig)
raw$religion_modifer <- unlist(lapply(tmp_relig_split,
function(x)
if(length(x) > 1)
paste(x[-1], collapse = "_") else
"missing"))
raw$religion_modifer <- resort_lvl(raw$religion_modifer)
###################################################################
## Same for sign
raw$sign <- gsub("’", "", raw$sign)
tmp_sign_split <- strsplit(raw$sign, split = " ")
tmp_sign <- unlist(lapply(tmp_sign_split, function(x) x[1]))
sign_lvl <- sort(unique(tmp_sign))
tmp_sign[tmp_sign == ""] <- "missing"
tmp_sign[is.na(tmp_sign)] <- "missing"
raw$sign <- resort_lvl(tmp_sign)
raw$sign_modifer <- unlist(lapply(tmp_sign_split,
function(x)
if(length(x) > 1)
paste(x[-1], collapse = "_") else
"missing"))
raw$sign_modifer <- resort_lvl(raw$sign_modifer)
###################################################################
## They are allowed to list multiple languages so we will pre-split
## these into dummy variables since they might have multiple choices.
## Also, "c++" and "lisp" !
tmp_speaks <- gsub("(", "", raw$speaks, fixed = TRUE)
tmp_speaks <- gsub(")", "", tmp_speaks, fixed = TRUE)
tmp_speaks <- gsub("c++", "cpp", tmp_speaks, fixed = TRUE)
tmp_speaks_split <- strsplit(tmp_speaks, split = ",")
tmp_speaks_split <- lapply(tmp_speaks_split,
function(x) gsub("^ ", "", x))
tmp_speaks_split <- lapply(tmp_speaks_split,
function(x) gsub(" ", "_", x))
speaks_values <- sort(unique(unlist(tmp_speaks_split)))
# tmp_speaks <- unlist(lapply(tmp_speaks_split, paste, collapse = ",", sep = ""))
for(i in speaks_values)
raw[, i] <- ifelse(unlist(lapply(tmp_speaks_split, function(x, sp) any(x == sp), sp = i)), 1, 0)
raw$speaks <- NULL
###################################################################
## Similaly, ethnicity is pre-split into dummy variables
tmp_eth <- gsub(", ", ",", raw$ethnicity)
tmp_eth <- gsub("/ ", "", tmp_eth)
tmp_eth <- gsub(" ", "_", tmp_eth)
tmp_eth_split <- strsplit(tmp_eth, split = ",")
eth_lvl <- sort(unique(unlist(tmp_eth_split)))
for(i in eth_lvl)
raw[, i] <- ifelse(unlist(lapply(tmp_speaks_split, function(x, eth) any(x == eth), eth = i)), 1, 0)
raw$ethnicity <- NULL
###################################################################
## There are very few missing values for continuous fields so
## remove them and convert the job field to the outcome.
okc <- raw[complete.cases(raw),]
okc$Class <- factor(ifelse(grepl("(computer)|(science)", okc$job), "stem", "other"),
levels = c("stem", "other"))
okc <- okc[okc$job != "missing",]
okc$job <- NULL
table(okc$Class)/nrow(okc)
save(okc, file = "okc.RData")