# ==============================
# LOAD LIBRARIES
# ==============================
library(tidyverse)
library(readxl)
library(ggplot2)
library(dplyr)

# ==============================
# LOAD DATA
# ==============================
data <- read_excel("financial_data.xlsx")

# Clean column names
names(data) <- make.names(names(data))

# Replace missing codes (if any)
data[data == 999] <- NA

# ==============================
# RECODING VARIABLES
# ==============================

# Sex (Male=1, Female=2)
data$sex_num <- ifelse(data$Sex == "Male", 1, 2)

# Marital Status
data$marital_num <- ifelse(data$Marital.Status == "Single", 1, 2)

# Yes/No variables
data$stock_num <- ifelse(data$Stock.Ownership == "Yes", 1, 0)
data$retire_num <- ifelse(data$Retirement.Plan == "Yes", 1, 0)
data$home_num <- ifelse(data$Home.Owner == "Yes", 1, 0)
data$promo_num <- ifelse(data$Promotion == "Yes", 1, 0)
data$trade_num <- ifelse(data$Trade.Last.Year == "Yes", 1, 0)
data$internet_num <- ifelse(data$Internet == "Yes", 1, 0)

# Education binary
data$edu_binary <- ifelse(data$Education %in% c("High School", "Less than High School"), 0, 1)

# ==============================
# DESCRIPTIVE STATISTICS
# ==============================

# Categorical frequencies
table(data$sex_num)
prop.table(table(data$sex_num))*100

table(data$marital_num)
prop.table(table(data$marital_num))*100

table(data$stock_num)
prop.table(table(data$stock_num))*100

# Continuous variables
summary(data$Age)
summary(data$Income)
summary(data$Kids)

# ==============================
# GRAPH 1: INCOME BY EDUCATION
# ==============================

plot1 <- data %>%
  group_by(Education) %>%
  summarise(mean_income = mean(Income, na.rm=TRUE))

ggplot(plot1, aes(x=Education, y=mean_income)) +
  geom_bar(stat="identity") +
  labs(title="Mean Family Income by Education Level",
       x="Education Level",
       y="Mean Income")

# ==============================
# GRAPH 2: AGE VS INCOME
# ==============================

ggplot(data, aes(x=Age, y=Income)) +
  geom_point() +
  labs(title="Relationship Between Age and Family Income",
       x="Age",
       y="Family Income")

# ==============================
# T-TEST (INCOME BY SEX)
# ==============================

t_test <- t.test(Income ~ sex_num, data=data)
print(t_test)

# ==============================
# CHI-SQUARE TESTS
# ==============================

# Retirement Plan vs Retirement Perception (q4)
chi1 <- chisq.test(table(data$retire_num, data$q4))
print(chi1)

# Promotion vs Job Security (q3)
chi2 <- chisq.test(table(data$promo_num, data$q3))
print(chi2)

# Home Ownership vs Education Perception (q4)
chi3 <- chisq.test(table(data$home_num, data$q4))
print(chi3)

# ==============================
# LOGISTIC REGRESSION
# ==============================

model <- glm(stock_num ~ Income + Age + Kids + edu_binary,
             data=data,
             family=binomial)

summary(model)

# ==============================
# SAVE OUTPUTS
# ==============================

dir.create("R_outputs", showWarnings = FALSE)

capture.output(t_test, file="R_outputs/t_test_income.txt")

capture.output(chi1, file="R_outputs/chi_retirement.txt")
capture.output(chi2, file="R_outputs/chi_promotion.txt")
capture.output(chi3, file="R_outputs/chi_homeowner.txt")

capture.output(summary(model), file="R_outputs/logistic_regression.txt")

print("ALL ANALYSIS COMPLETED SUCCESSFULLY")