-
Notifications
You must be signed in to change notification settings - Fork 0
Expand file tree
/
Copy pathCourseProject.R
More file actions
163 lines (131 loc) · 8.58 KB
/
CourseProject.R
File metadata and controls
163 lines (131 loc) · 8.58 KB
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
############################
# R Code for Final Project #
############################
# Install Packages
install.packages("ggplot2")
install.packages("dplyr")
install.packages("naivebayes")
install.packages("party")
install.packages("randomForest")
library(ggplot2)
library(dplyr)
library(naivebayes)
library(randomForest)
library(party)
# set your working directory to the current file directory
setwd(dirname(rstudioapi::getActiveDocumentContext()$path))
# Read DataSet from CSV and add column headers
dataset <- read.csv("Data-Files/Census-Project.data", header = FALSE)
colnames(dataset) <- c("Age", "WorkClass", "fnlwgt", "Education", "EducationNum", "MaritalStatus", "Occupation", "Relationship", "Race", "Sex", "CapitalGain", "CapitalLoss", "HoursPerWeek", "NativeCountry", "Income")
# Replace ? NA with and remove lines missing values
dataset[dataset == "?"] <- NA
#dataset$Income <- replace(dataset$Income, dataset$Income=='<=50K', 0)
#dataset$Income <- replace(dataset$Income, dataset$Income=='>50K', 1)
dataset$Income <- as.factor(dataset$Income)
initialSet <- na.omit(dataset)
# Consolidating Data Attributes to be used in analysis
## Education
LessThanHS <- c("Preschool", "1st-4th", "5th-6th", "7th-8th")
initialSet$Education[initialSet$Education %in% LessThanHS] <- "LessThanHS"
someHS <- c("9th", "10th", "11th", "12th")
initialSet$Education[initialSet$Education %in% someHS] <- "Some-HS"
Assoc <- c("Assoc-acdm", "Assoc-voc")
initialSet$Education[initialSet$Education %in% Assoc] <- "Assoc"
initialSet$Education <- as.factor(initialSet$Education)
initialSet$Education <- factor(initialSet$Education, levels = c("LessThanHS", "Some-HS", "HS-grad", "Some-college", "Assoc", "Prof-school", "Bachelors", "Masters", "Doctorate"))
## MaritalStatus
Married <- c("Married-civ-spouse", "Married-AF-spouse", "Married-spouse-absent")
initialSet$MaritalStatus[initialSet$MaritalStatus %in% Married] <- "Married"
initialSet$MaritalStatus <- as.factor(initialSet$MaritalStatus)
## Work Class
selfempClass <- c("Self-emp-inc", "Self-emp-not-inc")
initialSet$WorkClass[initialSet$WorkClass %in% selfempClass] <- "Self-emp"
govClass <- c("State-gov", "Federal-gov", "Local-gov")
initialSet$WorkClass[initialSet$WorkClass %in% govClass] <- "Government"
initialSet$WorkClass <- as.factor(initialSet$WorkClass)
## Occupation
initialSet$Occupation <- as.factor(initialSet$Occupation)
# Summarize Data
table(initialSet$WorkClass)
table(initialSet$Occupation)
table(initialSet$Education)
table(initialSet$MaritalStatus)
table(initialSet$Race)
table(initialSet$Sex)
table(initialSet$NativeCountry)
# Plotting
ggplot(initialSet, aes(y=Age, x=Income, fill=Income)) + geom_boxplot() + ggtitle("Age Distribution in Each Income Class")
ggplot(initialSet, aes(x=Income, y=Age, fill=Income)) + geom_violin(trim = FALSE, adjust=2) + geom_boxplot(width=0.1, color="grey", fill="white")
ggplot(initialSet, aes(x=Race, fill=Race)) + geom_bar(position="stack", stat="count") + ggtitle("Distribution of Races in Dataset")
ggplot(initialSet, aes(x=Sex, fill=Sex)) + geom_bar(position="stack", stat="count") + ggtitle("Distribution of Sexes in Dataset")
ggplot(initialSet, aes(x=Education, fill=Education)) + geom_bar(position="stack", stat="count") + ggtitle("Distribution of Education in Dataset")
ggplot(initialSet, aes(x=MaritalStatus, fill=MaritalStatus)) + geom_bar(position="stack", stat="count") + ggtitle("Distribution of Marital Status in Dataset")
ggplot(initialSet, aes(x=WorkClass, fill=WorkClass)) + geom_bar(position="stack", stat="count") + ggtitle("Distribution of Work Class in Dataset")
ggplot(initialSet, aes(x=Occupation, fill=Occupation)) + geom_bar(position="stack", stat="count") + ggtitle("Distribution of Occupation in Dataset")
ggplot(initialSet, aes(x=WorkClass, fill=Occupation)) + geom_bar(position="stack", stat="count", color="black") + ggtitle("Distribution of Occupation Per Work Class")
ggplot(initialSet, aes(x=Income, fill=Race)) + geom_bar(position="stack", stat="count") + ggtitle("Distribution of Races in Dataset based on Income Class")
ggplot(initialSet, aes(x=Race, fill=Income)) + geom_bar(position="dodge", stat="count") + ggtitle("Distribution of Races in Dataset based on Income Class")
ggplot(initialSet, aes(x=Income, fill=Sex)) + geom_bar(position="stack", stat="count") + ggtitle("Distribution of Sexes in Dataset based on Income Class")
ggplot(initialSet, aes(x=Sex, fill=Income)) + geom_bar(position="dodge", stat="count") + ggtitle("Distribution of Sexes in Dataset based on Income Class")
ggplot(initialSet, aes(x=Income, fill=MaritalStatus)) + geom_bar(position="stack", stat="count") + ggtitle("Distribution of Marital Status per Income Class")
ggplot(initialSet, aes(x=MaritalStatus, fill=Income)) + geom_bar(position="dodge", stat="count") + ggtitle("Distribution of Marital Status per Income Class")
ggplot(initialSet, aes(x=Income, fill=WorkClass)) + geom_bar(position="stack", stat="count") + ggtitle("Distribution of Work-Classes per Income Class")
ggplot(initialSet, aes(x=WorkClass, fill=Income)) + geom_bar(position="dodge", stat="count") + ggtitle("Distribution of Work-Classes per Income Class")
ggplot(initialSet, aes(x=Income, fill=Occupation)) + geom_bar(position="stack", stat="count") + ggtitle("Distribution of Occupations per Income Class")
ggplot(initialSet, aes(x=Occupation, fill=Income)) + geom_bar(position="dodge", stat="count") + ggtitle("Distribution of Occupations per Income Class")
ggplot(initialSet, aes(x=Income, fill=Education)) + geom_bar(position="stack", stat="count") + ggtitle("Distribution of Education per Income Class")
ggplot(initialSet, aes(x=Education, fill=Income)) + geom_bar(position="dodge", stat="count") + ggtitle("Distribution of Education per Income Class")
#ggplot(initialSet, aes(x=Income, fill=Relationship)) + geom_bar(position="stack", stat="count") + ggtitle("Distribution of Familial Relationship Status per Income Class")
# This one is kind of insane - we need a better way of showing: ggplot(initialSet, aes(x=Income, fill=NativeCountry)) + geom_bar(position="stack", stat="count") + ggtitle("Distribution of Country of Origin per Income Class")
# I tried this one as boxplot and scatter plot and it just looks jank - ggplot(initialSet, aes(x=Income, y=HoursPerWeek, fill=Income)) + geom_boxplot()
# Transform Categorical Data to Numeric Data
## Created a test Dataset to try this on
analysisSet <- initialSet
### analysisSet$NativeCountry[analysisSet$NativeCountry !="United-States"] <- "Outside-US"
analysisSet$NativeCountry[analysisSet$NativeCountry !="United-States"] <- 1
analysisSet$NativeCountry[analysisSet$NativeCountry == "United-States"] <- 0
analysisSet$Sex[analysisSet$Sex == "Male"] <- 0.0
analysisSet$Sex[analysisSet$Sex == "Female"] <- 1.0
analysisSet$Sex <- as.integer(analysisSet$Sex)
analysisSet$Education <- sapply(analysisSet$Education, unclass)
maxEducation <- as.integer(max(analysisSet$Education))
analysisSet <- transform(analysisSet, Education = Education / maxEducation, 1)
maxAge <- max(analysisSet$Age)
analysisSet <- transform(analysisSet, Age = Age / maxAge)
analysisSet$MaritalStatus <- sapply(analysisSet$MaritalStatus, unclass)
maxMaritalStatus <- as.integer(max(analysisSet$MaritalStatus))
analysisSet <- transform(analysisSet, MaritalStatus = MaritalStatus/maxMaritalStatus, 1)
analysisSet$WorkClass <- sapply(analysisSet$WorkClass, unclass)
maxWorkClass <- as.integer(max(analysisSet$WorkClass))
analysisSet <- transform(analysisSet, WorkClass = WorkClass/maxWorkClass, 1)
analysisSet$Occupation <- sapply(analysisSet$Occupation, unclass)
maxOccupation <- as.integer(max(analysisSet$Occupation))
analysisSet <- transform(analysisSet, Occupation = Occupation/maxOccupation, 1)
# Decision Tree
index <- sample(2, nrow(analysisSet), replace=TRUE, prob=c(0.8, 0.2))
TrainingSet <- analysisSet[index==1,]
TestSet <- analysisSet[index==2,]
Formula <- Income ~ Age + WorkClass + Education + MaritalStatus + Occupation + Sex
cTree <- ctree(Formula, data=TrainingSet)
train_predict <- predict(cTree, TrainingSet)
table(train_predict, TrainingSet$Income)
test_predict <- predict(cTree, TestSet)
table(test_predict, TestSet$Income)
# Random Forest
rf <- randomForest(formula=Formula, data=TrainingSet)
plot(rf)
train_predict <- predict(rf, TrainingSet)
t1 <- table(train_predict, TrainingSet$Income)
1 - sum(diag(t1)) / sum(t1)
test_predict <- predict(rf, TestSet)
t2 <- table(test_predict, TestSet$Income)
1 - sum(diag(t2)) / sum(t2)
# Naive Bayes
nb <- naive_bayes(formula=Formula, data=TrainingSet)
plot(nb)
train_predict <- predict(nb, TrainingSet)
t1 <- table(train_predict, TrainingSet$Income)
1 - sum(diag(t1)) / sum(t1)
test_predict <- predict(nb, TestSet)
t2 <- table(test_predict, TestSet$Income)
1 - sum(diag(t2)) / sum(t2)