Main page.Rmd

---
title: "Religion in South Korea"
author: "Grace Kim"
date: "11/13/2019"
output: html_document
---

This was a practice document for the initial creation of all the various graphics in this project. This contains many of the bar graphs for Gender, Age, and the world v.s. south korea bargraph.

```{r setup, include=FALSE}
knitr::opts_chunk$set(echo = FALSE)

# Need all of the libraries!

library(tidyverse)
library(readxl)
library(janitor)
library(hrbrthemes)
library(openxlsx)
library(stringr)

```

```{r}
# 2015 south Korea census data

x <- read_xlsx("Korea/raw-data/2015_Religions_Korea.xlsx") %>% 
  clean_names()

#fixes merged column problem from excel for the dataset

test<- x %>% fill(gender) %>% fill(korean_region)

# Data from world peabody research center, has numbers of every nations various religious
# populations data needed to set the various column attributes as its a csv

world<- read_csv("Korea/raw-data/World_Religion_Data.csv", col_types = cols(
  Country = col_character(),
  Christian = col_double(),
  Muslim = col_double(),
  Unaffiliated = col_double(),
  Hindu = col_double(),
  Buddhist = col_double(),
  Folk = col_double(),
  Other = col_double(),
  Jewish = col_double()
)) %>% 
  clean_names()

```


```{r Gender circular bar plot}

#trying to filter data for only the total number of korean data
gender_data <- x %>% 
  filter(korean_region == "All of Korea") %>% 
  filter(gender != "Korea - Total") %>% 
  filter(age == "Total")

#needs to be a data frame in order to use gather

gender_data <- data.frame(gender_data)

#gather flips the rows and columns

gender_data_vertical <- gather(gender_data)

# take only the columns we need for the grouped plot

gender_data_vertical <- gender_data_vertical[11:30,] 

# formatting for the various column names

gender_data_vertical<- gender_data_vertical %>% 
  mutate(gender = ifelse(row_number()%%2 == 0,
                               "Female",
                               "Male")) %>% 
  mutate(religion = ifelse(row_number()%%2 == 0,
                               str_c(key, "F"),
                               str_c(key, "M")))

# add the necessary row numbers needed

gender_data_vertical <- transform(gender_data_vertical, value = as.numeric(value)) 

# arranged from desc

gender_data_vertical<- gender_data_vertical%>% 
  arrange(desc(value))

# plotted the percentages of the different gender groups that practice each religion
# filled the entire chart so it could be a percentage based approximation so we could
# compare each religious group on a normalized scale

ggplot(gender_data_vertical,aes(x = reorder(as.factor(religion),value), y = value, fill = gender)) +
  geom_bar(stat = "identity") +
  coord_flip() +
  scale_fill_manual(values = c("#FF5733","#15C9C1")) +
  labs(title = "Male and Female Religous Data", x = "Religions", y = "Percentage Practiced by Population")


gender <- ggplot(gender_data_vertical, aes(x = reorder(as.factor(religion),value), y = value, fill = gender)) +       
  # Note that id is a factor. If x is numeric, there is some space between the first bar
  
  # This add the bars with a blue color
  geom_bar(stat="identity") +
  
  scale_y_log10()+
  theme_minimal()+
  theme(
    axis.title = element_blank(),
    panel.grid = element_blank(),
    plot.margin = unit(rep(-1,4), "cm") 
  ) +
  # This makes the coordinate polar instead of cartesian.
  coord_polar(start = 0)
gender
```

```{r Age line graphs}
# data cleaning for age to get only the total populations of each age,
# needed to make sure age was a numeric variable as well by only taking the first
# two characters of the row names

age_data <- x %>% 
  filter(korean_region == "All of Korea") %>% 
  filter(age != "Total") %>% 
  filter(gender == "Korea - Total") %>% 
  select(age, religious_total) %>% 
  mutate(age_number = as.numeric(substr(age,1,2))) %>% 
  mutate(age2 = as.numeric(age_number)^2)

#only so i can see the numbers
quadratic.model <- lm(data = age_data, religious_total ~ age_number + age2)
summary(quadratic.model)

# had a age plot with a quadratic modeled on top of it so that we can see the trends
# as the religious populations move within generations

ggplot(age_data, aes(x = as.numeric(age_number), y = religious_total, fill = age_number)) +
  geom_bar(stat = "identity")+
  stat_smooth(method = "lm", formula = y ~ x + I(x^2), size = 1)
```


```{r World Religions}
# summarized all the percentages of the nations around the world

world_summary<- world %>% 
  summarize(christian = mean(as.numeric(christian)),
            muslim = mean(as.numeric(muslim)),
            unaffiliated = mean(as.numeric(unaffiliated)),
            hindu = mean(as.numeric(hindu)),
            buddhist = mean(as.numeric(buddhist)),
            folk = mean(as.numeric(folk)),
            other = mean(as.numeric(other)),
            jewish = mean(as.numeric(jewish))) 

# took the korea totals from the world dataset

korea_totals <- world %>% 
  filter(country == "South Korea") %>% 
  select(-matches("country"))

# merged the two world summary and korea totals data together

world_korea_summary <- rbind(data.frame(world_summary),data.frame(korea_totals))

# changed rows and columns because of the grouped bar plot format we needed

vertical_world <- gather(world_korea_summary)%>% 
  mutate(country = ifelse(row_number()%%2 == 1, "World","South Korea"))

# grouped barplot to show the various religious follower populations between south Korea
# v.s. the rest of the world

ggplot(vertical_world, aes(x = key, y=value, fill = country))+ 
  geom_bar(position="dodge", stat="identity")+
  labs(x = "Religion",
       y = "Percentages of Religious Followers",
       fill = "",
       title = "The Religions of the World vs. South Korea")

```

summarize(christianity = mean(as.numeric(christian))) %>% DONT LOOK DOWN HERE:



```{r OLD data cleaning}

korean_migration_data<- read_excel("Korea/raw-data/LongTerm_International_Migration_by_Sex_and_Age_Nationals__Foreigners__20191005095057.xlsx") %>% clean_names()

```

```{r OLD Bar Graph}
# Need to sort out my data first in order to make the grouped barchart

expanded_korean <- korean_migration_data %>% 
  filter(by_gender != "Total") %>%
  filter(by_age != "unknown", by_age!="Total") %>%
  
  # Filtered out total and unknowns since that was the duplicate of the combined data
  
  group_by(by_age, by_gender) %>% 
  summarize(incoming_migrants_total = sum(incoming_migrants),
            outcoming_migrants_total = sum(outcoming_migrants)) %>% 
  
  # Summarized the data that I had from the incoming and leaving migrants, needed to 
  # combine this data however into one column in order for me to make a grouped plot
  
  mutate(freq = 2)

  # made a column that would allow me to duplicate each row

expanded_korean
expanded_korean <- expanded_korean[rep(row.names(expanded_korean),
                                       expanded_korean$freq), 1:4]

# Duplicated each row of the the expanded Korean dataset so I could have the incoming and 
# leaving migrant numbers all in one column


# Made several rows that could contain both numbers for leaving and outcoming migrants
# had a identifier row called migrant_type and numbers in migrant_num

barplot_data<- expanded_korean %>% 
  mutate(migrant_type = ifelse(row_number()%%2 == 0,
                               "Incoming",
                               "Leaving")) %>% 
  mutate(migrant_num = ifelse(migrant_type == "Leaving",
                              outcoming_migrants_total,
                              incoming_migrants_total))


# Time to make the graph of our grouped barplot!
# needed to set the groups by migrant type and by age; also set the scale of y axis so 
# that there wasn't weird scientific notation. Also split up by gender


ggplot(barplot_data, aes(x = by_age, y=migrant_num, fill = migrant_type))+ 
  geom_bar(position="dodge", stat="identity")+
  facet_grid(~by_gender) +
  labs(x = "",
       y = "#'s of incoming/leaving migrants",
       fill = "Migrant Type",
       title = "Migrants of South Korea")+
  scale_y_continuous(labels = scales::comma)
```