# Printing nicely in R Markdown (option 1)
flextable::flextable(dataDictionary, cwidth = 2) # Printing the dataDictionary table using the flextable R package

# Creating the region variable
# Used chatGPT to help list all states with their respective regions
census_data <- census_data %>%
  mutate(
    region = case_when(
      str_detect(county_state, "California|Oregon|Washington|Alaska|Hawaii|Arizona|Colorado|Idaho|Montana|Nevada|New Mexico|Utah|Wyoming") ~ "West", # Detecting West regional states
      str_detect(county_state, "North Dakota|South Dakota|Nebraska|Kansas|Minnesota|Iowa|Missouri|Wisconsin|Illinois|Michigan|Indiana|Ohio") ~ "Midwest", # Detecting Midwest regional states
      str_detect(county_state, "Texas|Oklahoma|Arkansas|Louisiana|Mississippi|Alabama|Florida|Georgia|South Carolina|North Carolina|Tennessee|Kentucky|Virginia|West Virginia|Maryland|Delaware|District of Columbia|Puerto Rico") ~ "South", # Detecting Southern regional states
      str_detect(county_state, "Maine|New Hampshire|Vermont|Massachusetts|Rhode Island|Connecticut|New York|New Jersey|Pennsylvania") ~ "Northeast", # Detecting Northeast regional states
      TRUE ~ NA_character_
    )
  )

# Creating summary table
regions <- census_data %>%
  group_by(region) %>%
  summarize(Unique_States = n_distinct(county_state), .groups = "drop")

# Display the summary table
flextable::flextable(regions)

# Creating the combined_other_race variable
census_data <- census_data %>% 
  mutate(
    combined_other_race = prop_hawaiin_islander + prop_other_race + prop_multi_racial
  )

# Calculating the average total percentage of race by region
raceByRegion <- census_data %>%
  group_by(region) %>%
  summarize(
    Average_White = mean(prop_white),
    Average_Black = mean(prop_black),
    Average_Native = mean(prop_native),
    Average_Asian = mean(prop_asian),
    Average_Other = mean(combined_other_race),
    .groups = "drop"
  )

# Display the table
flextable::flextable(raceByRegion)

# Creating the expenses variable
expenses <- census_data %>% 
  group_by(year) %>%
  summarize(
    income_after_home_expenses = label_currency(mean((median_income - (12 * (median_monthly_home_cost))))), # Calculates the average median anual income after home costs
    income_after_rent_expenses = label_currency(mean((median_income - (12 * (median_monthly_rent_cost))))) # Calculates the average median anual income after rent costs
    )
# Having the year displayed without commas
expenses$year <- as.character(expenses$year)

# Displaying the table
flextable::flextable(expenses, cwidth = 2.5)

# Used chatgpt to help with ifelse statements
# Creating the freq_table variable
freq_table <- census_data %>%
  mutate(Gender_Group = ifelse(prop_female > 0.5, "Female", "Male"), # Finds female or male dominated states
         Poverty_Group = ifelse(prop_poverty > 0.145, "High Poverty", "Low Poverty")) %>% # Finds if they are above or below the national average poverty percentage
  group_by(Gender_Group, Poverty_Group) %>%
  summarize(Frequency = n(), 
            .groups = "drop")

# Displaying the frequency table using flextable
flextable::flextable(freq_table, cwidth = 1.5)

# Creating the variable avg_costs
avg_costs <- census_data %>%
  pivot_longer(cols = c(median_monthly_home_cost, median_monthly_rent_cost), # Used chatgpt to help combine these two plot together
               names_to = "cost_type", 
               values_to = "value") %>%
  group_by(year, cost_type) %>%
  summarize(avg_value = mean(value), 
            .groups = "drop")

# Creating the variable housingLineChart to plot home and rent expenses
housingLineChart <- avg_costs %>%
  ggplot(aes(x = year, 
             y = avg_value, 
             color = cost_type, 
             group = cost_type)) +
  geom_line(linewidth = 1) +
  labs(title = "Median Monthly Housing Costs Over the Years in the US",
       x = "Year",
       y = "Average Median Monthly Cost",
       caption = "Data source: census_data_state_2008-2021.csv") +
  scale_color_colorblind() + 
  theme_bw() + 
  theme(legend.position = "bottom") + 
  scale_y_continuous(labels = scales::dollar_format(prefix = "$")) # Formats the y-axis to use a currency format

# Displays the graph
housingLineChart

# Creating the combined_degree_holders variable combining the percentages of all degree holding Americans
census_data <- census_data %>%
  mutate(
    combined_degree_holders = prop_associates + prop_bachelors + prop_masters + prop_professional + prop_doctoral
  )

# Creating the combined_non_degree_holders variable combining the percentages of all non-degree holding Americans
census_data <- census_data %>%
  mutate(
    combined_non_degree_holders = prop_highschool + prop_GED + prop_some_college + prop_college_no_degree
  )

# Combining degree holders and non-degree holders
long_data <- census_data %>%
  select(year, combined_degree_holders, combined_non_degree_holders) %>%
  pivot_longer(cols = c(combined_degree_holders, combined_non_degree_holders),
               names_to = "Education_Level",
               values_to = "Percentage") %>%
  group_by(year, Education_Level) %>%
  summarize(avg_perc = mean(Percentage), 
            .groups = "drop")

# Plotting the data
degreeLineChart <- ggplot(long_data, aes(x = year, 
                                           y = avg_perc, 
                                           color = `Education_Level`, 
                                           group = `Education_Level`)) +
  geom_line(linewidth = 1) +
  labs(title = "Education Level Over the Years in the US",
       x = "Year",
       y = "Combined Percentage",
       caption = "Data source: census_data_state_2008-2021.csv") +
  scale_color_colorblind() +
  theme_bw() +
  theme(legend.position = "bottom")

# Plotting the graph
degreeLineChart

# Calculate average median income per year and region
average_income_per_year <- census_data %>%
  group_by(year, region) %>%
  summarize(avg_median_income = mean(median_income), .groups = "drop")

# Plotting the data
incomeLineChart <- average_income_per_year %>% 
  ggplot(aes(x = year, 
             y = avg_median_income, 
             color = region, 
             group = region)) +
  geom_line(linewidth = 1) +
  labs(title = "Average Median Income by Region",
       x = "Year",
       y = "Average Median Income",
       caption = "Data source: census_data_state_2008-2021.csv") +
  scale_color_colorblind() +
  theme_classic() +
  theme(legend.position = "bottom") + 
  scale_y_continuous(labels = scales::dollar_format(prefix = "$")) + # Formatting to show currency in the y-axis
  guides(color = guide_legend(title = "Region"))

# Displaying the graph
incomeLineChart

# Calculate average median home and rental costs per year and region
avg_costs_by_region <- census_data %>%
  pivot_longer(cols = c(median_monthly_home_cost, median_monthly_rent_cost), 
               names_to = "cost_type", 
               values_to = "value") %>%
  group_by(year, region) %>%
  summarize(avg_value = mean(value), .groups = "drop")

# Plotting the data
livingCostsLineChart <- avg_costs_by_region %>%
  ggplot(aes(x = year, 
             y = avg_value, 
             color = region, 
             group = region)) +
  geom_line(linewidth = 1) +
  labs(title = "Average Median Monthly Housing Costs by Region",
       x = "Year",
       y = "Average Median Monthly Cost",
       caption = "Data source: census_data_state_2008-2021.csv") +
  scale_color_colorblind() + 
  theme_classic() + 
  theme(legend.position = "bottom") + 
  scale_y_continuous(labels = scales::dollar_format(prefix = "$")) + # Formatting to show currency in the y-axis
  guides(color = guide_legend(title = "Region"))

# Displaying the graph
livingCostsLineChart

# Creating the incomeDegree variable
incomeDegree <- census_data %>%
  ggplot(aes(
    x = combined_degree_holders,
    y = median_income,
    color = region, # Coloring by region
    size = population # Setting the size to population size
  )) +
  geom_point(alpha = 0.2) +
  scale_color_colorblind() +
  labs(
    title = "Median Income vs. Percentage of Degree Holders by Region",
    x = "Combined Degree Holders (%)",
    y = "Median Income",
    caption = "Data source: census_data_state_2008-2021.csv", 
    color = NULL
  ) +
  facet_grid(. ~ region) + # Creating the four individual graphs based on region
  theme(legend.position = "bottom") +  
  geom_smooth(method = "lm", se = FALSE, size = 1) + 
  guides(
    size = guide_legend(title = "Population Size (Millions of People)"), 
    color = FALSE
  ) + 
  scale_y_continuous(labels = scales::dollar_format(prefix = "$")) + 
  scale_size_continuous(labels = scales::comma_format(scale = 1e-6, suffix = "M")) # Changing the legend to show the number in millions of people rather than in scientific notation

# Displaying the graph
incomeDegree

# Calculating average values per year
avg_data <- census_data %>%
  group_by(year) %>%
  summarize(
    avg_percent_poverty = mean(prop_poverty),
    avg_monthly_housing_costs = mean((median_monthly_home_cost + median_monthly_rent_cost) / 2),
    avg_median_income = mean(median_income)
  )

# Calculating percentage changes
avg_data_percent_change <- avg_data %>% # Used chatGPT to help with this
  mutate(
    median_income_change = (avg_median_income / lag(avg_median_income) - 1) * 100,
    housing_costs_change = ((avg_monthly_housing_costs + lag(avg_monthly_housing_costs)) / (2 * lag(avg_monthly_housing_costs)) - 1) * 100,
    percent_poverty_change = (avg_percent_poverty / lag(avg_percent_poverty) - 1) * 100
  )

# Reshaping data for plotting
avg_data_change_long <- avg_data_percent_change %>% 
  pivot_longer(cols = c(median_income_change, housing_costs_change, percent_poverty_change),
               names_to = "variable",
               values_to = "percent_change")

# Plotting the percent change
percentChange <- avg_data_change_long %>%
  ggplot(aes(x = year, y = percent_change, fill = variable)) +
  geom_col(position = "dodge", width = 0.7, color = "white") +
  labs(title = "Percentage Change in Median Income, Housing Costs, and Poverty Over the Years",
       x = "Year",
       y = "Percentage Change",
       fill = "Variable",
       caption = "Data source: census_data_state_2008-2021.csv") +
  scale_fill_manual(values = c("skyblue", "orange", "lightgreen")) + # Changing the color of each bar
  theme_minimal() +
  theme(axis.text.x = element_text(angle = 45, hjust = 1))

percentChange