RViz_Happiness.Rmd

---
title: "Happiness"
subtitle: "A Journey Around the Globe."
author: "Dustin Pacholleck & Adnan Sevinc"
date: "2023-01-15"
output: 
  ioslides_presentation:
    widescreen: true
    logo: img/wne_logo2017.png 
  
---

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



```{r settings_1, include=FALSE}
#install needed packages more easily with pacman
if (!require("pacman")) install.packages("pacman")

pacman::p_load(                                # Install or/and import libraries.
               "ggplot2",
               "tidyverse",
               "dplyr",
               "CGPfunctions",
               "rnaturalearth",
               "sf",
               "ggrepel",
               "ggthemes",
               "gridExtra",
               "cowplot",
               "ggforce",
               "GGally",
               "Rmisc",
               "grid",
               "viridis",
               "RColorBrewer",
               "gganimate",
               "gifski",
               "wbstats",
               "directlabels",
               "reshape2",
               "devtools",
               "ggstance",
               "patchwork",
               "aplot",
               "tidypaleo",
               "gapminder",
               "ggpmisc"
)


Sys.setlocale("LC_TIME", "English")

# Set the current path as directory
setwd(dirname(rstudioapi::getSourceEditorContext()$path))   

# Import dataset
data <- read.csv("data/DataForTable2.1.csv") 
# Create copy of dataset for future work
data_raw <- data %>% 
            select(-c(Negative.affect,Positive.affect,Confidence.in.national.government))  # Delete Columns

# Rename Columns
names(data_raw)<-c("Country",                       
               "Year", 
               "HappinessScore",
               "GDPPer",
               "SocialSupport",
               "LifeExpectancy",
               "Freedom",
               "Generosity",
               "Corruption")

# Import additional stats from World Bank
# Define indicators
my_indicators <- c(                         
  life_exp = "SP.DYN.LE00.IN", 
  gdp_capita ="NY.GDP.PCAP.CD", 
  pop = "SP.POP.TOTL"
)
# Query data
world_stats <- wb_data(my_indicators, start_date = min(data$year), end_date = max(data$year)) 

# Inspecting dataframe
data <- data[complete.cases(data),]  #only complete dataset
nrow(data)
 
# Drop the below columns
data$Negative.affect<-NULL     
data$Positive.affect<-NULL     
data$Confidence.in.national.government<-NULL

#Rename Columns
names(data)
names(data)<-c("Country",          
               "Year", 
               "HappinessScore",
               "GDPPer",
               "SocialSupport",
               "LifeExpectancy",
               "Freedom",
               "Generosity",
               "Corruption")

# Complete dataset with missing years
countries <- unique(data$Country)
years <- unique(data$Year)
all <- data.frame(expand.grid(countries,years))
names(all) <- c("Country", "Year")
data <- left_join(all,data, by = c("Country"="Country", "Year" = "Year"))

data <- left_join(data,world_stats, by = c("Country"="country", "Year" = "date"))

world <- ne_countries(scale = "medium", returnclass = "sf")

data <- left_join(data, world[c("name","economy","continent", "subregion")], by = c("Country"="name"))
```

## Happiness Around the World {.flexbox .vcenter}

```{r Map World, echo=FALSE}
# Filter to 2021
data_2021 <- data %>% 
  mutate(HappinessScore = round(HappinessScore,2)) %>% 
  filter(Year == 2021)

# Join spacial data with dataset
data_world <- full_join(data_2021, world, by = c("Country"="name","economy","continent", "subregion","geometry"))

# Create base world plot
theme_set(theme_bw())

gworld_base <- 
  ggplot(data = data_world) +
  coord_sf(expand = FALSE) +
  geom_sf(aes(geometry =  geometry, fill =  HappinessScore)) +
  scale_fill_gradient(low="red", high="green",na.value = "grey") 

# Create World Plot  
gworld <- 
  gworld_base +
  theme(legend.position = 'bottom',
        plot.title = element_text(size = 20, hjust = 0.5)) +
  labs(title = "Happiness Worldwide in 2021",
       caption = "Own creation. Data: World Happiness Report, World Bank")

gworld
```

## Distribution of Happiness {.flexbox .vcenter}

```{r Hist, echo=FALSE}
# Prep data
data_hist <- data %>% 
            drop_na(continent) %>%   # Drop na from continent 
            select(c("continent",    # Select specific Column
                    "Country", 
                     "Year", 
                     "HappinessScore",
                     "GDPPer",
                     "SocialSupport",
                     "LifeExpectancy",
                     "Freedom",
                     "Generosity",
                     "Corruption"))

# Generate plot
ggplot(data_hist %>% filter(continent != "Seven seas (open ocean)"), aes(x = HappinessScore, fill = continent)) +  # Filter out Open seas.
  geom_histogram(fill = "grey", alpha = .5) +
  geom_histogram(colour = "black") +
  facet_wrap(~ continent) + # separate by continent.
  guides(fill = FALSE)+                       
  labs(title = "Distribution of Happiness Score by Region",
       x = "",
       y = "",
       caption = "Own creation. Data: World Happiness Report, World Bank")+
  theme(plot.title = element_text(size = 15, hjust = 0.5))+
  theme_stata()   # Use stata theme
```

## Happiness Ranking  {.flexbox .vcenter}

```{r BarPlot2, echo=FALSE}
# Prepare data
data_barplot_country <-   data %>%   # Mean Happiness Score by Country
  select(c( "Country", 
            "HappinessScore"
  )) %>% 
  drop_na(Country) %>%  # Drop Na
  group_by(Country) %>% 
  summarise_all(mean , na.rm = TRUE) # Summarise all column by mean


data_barplot_country_top10 <-   data %>%   #Top 10 countries in terms of Happiness score
  select(c( "Country", 
            "HappinessScore",
  )) %>% 
  drop_na(Country) %>% 
  group_by(Country) %>% 
  summarise_all(mean , na.rm = TRUE) %>% 
  top_n(10) %>%                        # First 10
  mutate (Position = 'Top10') %>%      # Flag as Top 10
  arrange(desc(HappinessScore))        # Descending Order



data_barplot_country_bottom10 <-   data %>% #Bottom 10 countries in terms of Happiness score
  select(c( "Country", 
            "HappinessScore")) %>% 
  drop_na(Country) %>% 
  group_by(Country) %>% 
  summarise_all(mean , na.rm = TRUE) %>% 
  top_n(-10) %>%                              # Bottom 10
  mutate (Position = 'Bottom10') %>%            # Flag as Bottom 10
  arrange(HappinessScore)      # Ascending Descending Order


data_barplot_country_top_bottom_union <- 
  union_all(data_barplot_country_top10,data_barplot_country_bottom10) %>%  # Union Top and Bottom 10 Countries
  arrange(HappinessScore)


data_barplot_country_top_bottom_union = data %>%  # Right Join in order to get Continent
  select(continent,Country) %>% 
  unique()  %>% 
  right_join(data_barplot_country_top_bottom_union, by ="Country")


data_barplot_country_top_bottom_union$HappinessScore <- round(data_barplot_country_top_bottom_union$HappinessScore,3) # Round Happiness Score
data_barplot_country_top_bottom_union$Country <- reorder(data_barplot_country_top_bottom_union$Country, data_barplot_country_top_bottom_union$HappinessScore)  # Reorder Country
data_barplot_country_top_bottom_union$continent <- reorder(data_barplot_country_top_bottom_union$continent, data_barplot_country_top_bottom_union$HappinessScore) # Reorder Happiness Score

data_barplot_country_top_bottom_union[is.na(data_barplot_country_top_bottom_union$continent),"continent"] <- "Africa" # Missing Continet Africa fill out.

freq_table <- data_barplot_country_top_bottom_union %>%   # Frequency Table of Top 10 in terms of Continent.
  select(continent,Position) %>% 
  count() %>% 
  arrange(desc(Position), desc(freq))

names(freq_table)[3] = '# Countries'  # Rename as # Countries
names(freq_table)[1] = 'Continent'    # Renames as Continent

# Generate Plot
ggplot(data_barplot_country_top_bottom_union, aes(x = Country, y= HappinessScore, fill = Position)) +
  geom_bar(stat='identity') +
  coord_flip()+
  labs(title = "Top and Bottom 10 Countries by Average Happiness",
       y = 'Average Happiness Score',
       x = NULL,
       fill = '',
       caption = "Own creation. Data: World Happiness Report, World Bank") +
  theme_minimal() +
  theme(legend.position = 'bottom',
        plot.title = element_text(hjust = 0.5),
        axis.text.x=element_blank(),
        axis.ticks.x=element_blank()) +
  geom_text(aes(label = HappinessScore), vjust = .2,hjust = -.1, position = position_dodge(.5),
            size = 3) +
  scale_fill_manual(values = c('grey78', 'khaki')) +
  geom_hline(yintercept = mean(data$HappinessScore, na.rm = TRUE,show.legend = TRUE), linetype="dashed", color = "red") +
  annotate(geom = 'table',
           x=0,
           y=12,
           label=list(freq_table))
```

## Happiness in Europe {.flexbox .vcenter}

```{r Map Europe, echo=FALSE}
# Use world base plot for subplot
gworld_sub <- 
  gworld_base +
  geom_rect(xmin = -30, xmax = 50, ymin = 35, ymax = 70, # Place rectange at Europe
            fill = NA, colour = "black", size = 1.5) +
  theme(legend.position= 'none',
        axis.text = element_blank(),
        axis.ticks=element_blank(),
  )

# Create plot for Europe
geurope <- ggplot(data = data_world) +
  coord_sf(expand = FALSE) +
  geom_sf(aes(geometry =  geometry, fill =  HappinessScore)) +
  annotate(geom = "text", x = 19, y = 52, label = "Poland", # Add Poland as text 
           fontface = "italic", color = "black", size = 4) +
  scale_fill_gradient(low="red", high="green",na.value = "grey") +
  coord_sf(xlim = c(-30, 50), ylim = c(35, 70), expand = TRUE) +
  theme(legend.position = 'bottom',
        plot.title = element_text(size = 20, hjust = 0.5)) +
  labs(title = "Happiness Europe in 2021",
       caption = "Own creation. Data: World Happiness Report, World Bank")

# Generate plot for Europe with world subplot
ggdraw() +
  draw_plot(geurope,
            0, 0, 1, 1) +
  draw_plot(gworld_sub, 
            0.03, 0.25, 0.2, 0.2) # subplot in left corner
```

## Change of Happiness Ranking {.flexbox .vcenter}

```{r Happiness Ranking 2, echo=FALSE} 
# Data prep
european_union <- c("Austria","Belgium","Bulgaria","Croatia","Cyprus",                # List of European Countries.
                    "Czech Rep.","Denmark","Estonia","Finland","France",
                    "Germany","Greece","Hungary","Ireland","Italy","Latvia",
                    "Lithuania","Luxembourg","Malta","Netherlands","Poland",
                    "Portugal","Romania","Slovakia","Slovenia","Spain",
                    "Sweden","United Kingdom")


european_union_data <-                                                               # Extract Just European Countries
                     data_raw %>% 
                      filter(Country %in% european_union)

european_union_data_top10 <- european_union_data %>%                                # Extract Top 10 European Countries
                            group_by(Country) %>%
                            summarise_all(mean) %>%
                            arrange(desc(HappinessScore)) %>% 
                            head(10)


european_union_data_top10_PL <- european_union_data %>% 
                                filter(Country %in% c(european_union_data_top10$Country , "Poland")) %>%    # Filter date
                                filter(Year> 2009) %>% 
                                filter(Year<= 2021)

data1 <- european_union_data_top10_PL
data1$Year <- as.character(data1$Year)   # Year to character for newggslopegraph.
data1$HappinessScore <- round(data1$HappinessScore,2) # Round happiness to 2 decimal


# Generate Plot
newggslopegraph(data1, Year, HappinessScore, Country,
                Title = "Top 10 European Countries & Poland's Happiness Score",          
                SubTitle = "2010-2021",
                Caption = "Own creation. Data: World Happiness Report, World Bank",
                DataLabelPadding = 0.2,
                DataLabelLineSize = 0.5,
                DataLabelFillColor = "lightblue") 

```

## Drivers of Happiness {.flexbox .vcenter}

```{r Scatter plot, echo=FALSE}
# Data prep
data_Scatter <- data_raw %>%  
        mutate(PolandFlag = ifelse(Country == "Poland", "Poland", "Trend Line")) # Flag the Country whether is Poland or Others.

# Generate Plots
g1 <- ggplot(data_Scatter,aes(x = GDPPer, y = HappinessScore,color = PolandFlag)) + 
  geom_point(color = "#adaaaa")+
  geom_smooth(alpha = 0.2) +
  geom_rect(xmin = 9.977814, xmax = 10.4364, ymin = 5.646205, ymax = 6.242094, # To rectangle Poland
            fill = NA, colour = "red", size = 1)+
  scale_y_continuous(breaks = seq(0, 8, by = 0.5))+                # y scale start from 0 till 8 with 0.5 incremental
  scale_x_continuous(breaks = seq(0, 12, by = 0.5),
                     labels = paste0('$',                          # Paste $ sign to x scale.
                                     seq(0, 12, by = 0.5)))+
  labs(title = '',
       x = 'GDP Per capita' , y = 'HappinessScore')

g2 <- ggplot(data_Scatter,aes(x = SocialSupport, y = HappinessScore,color = PolandFlag)) + 
  geom_point(color = "#adaaaa") +
  geom_rect(xmin = 0.8634442, xmax = 0.9550653, ymin = 5.646205, ymax = 6.242094, 
            fill = NA, colour = "red", size = 1)+
  geom_smooth()+
  scale_y_continuous(breaks = seq(0, 8, by = 0.5))+
  scale_x_continuous(breaks = seq(0, 1, by = 0.1))+
  labs(title = '',
       x = 'Social Support' , y = 'HappinessScore')

g3 <- ggplot(data_Scatter,aes(x = LifeExpectancy, y = HappinessScore,color = PolandFlag)) + 
  geom_point(color = "#adaaaa") +
  geom_smooth()+
  geom_rect(xmin = 66.56, xmax = 69.05, ymin = 5.646205, ymax = 6.242094, 
            fill = NA, colour = "red", size = 1)+
  scale_y_continuous(breaks = seq(0, 8, by = 0.5))+
  scale_x_continuous(breaks = seq(10, 70, by = 10))+
  labs(title = '',
       x = 'Life Expectancy' , y = 'HappinessScore')

g4 <- ggplot(data_Scatter,aes(x = Freedom, y = HappinessScore,color = PolandFlag)) + 
  geom_point(color = "#adaaaa")+
  geom_smooth()+
  geom_rect(xmin = 0.7318056, xmax = 0.8828858, ymin = 5.646205, ymax = 6.242094, 
            fill = NA, colour = "red", size = 1)+
  scale_y_continuous(breaks = seq(0, 8, by = 0.5))+
  scale_x_continuous(breaks = seq(0, 1, by = 0.1))+
  labs(title = '',
       color = '',
       x = 'Freedom' , y = 'HappinessScore')

g4 <- g4 + theme(legend.position = "bottom")
# Get legend from last plot
legend <- get_legend(g4) 

# Title formatting
title <- textGrob("Correlation of Happiness Score",
                  gp = gpar(fontsize = 15, # font size
                            fontface = 2) # bold type
)

# Arrange plots
grid.arrange(arrangeGrob(g1 + theme(legend.position = "none"),  #We use grid.arrange to display 4 graphs.
                         g2 + theme(legend.position = "none"), 
                         g3 + theme(legend.position = "none"), 
                         g4 + theme(legend.position = "none"), 
                         ncol = 2), 
             legend, 
             nrow = 2, 
             top = title,  
             heights = c(10, 1) 
) 
```

<font size="1"> *Own creation. Data: World Happiness Report, World Bank* </font>

## Happiness and GDP {.flexbox .vcenter}

```{r Bubble plot, include=FALSE}

## Search for countries with biggest difference in score for highlighting
# focus <- unique(data$Country %>% filter(Year > 2010 & continent == 'Europe))
# inspect <- data_bubble[c("Country", 
#                          "Year", 
#                          "HappinessScore",
#                          "GDPPer",
#                          "SocialSupport",
#                          "LifeExpectancy",
#                          "Freedom",
#                          "Generosity",
#                          "Corruption")] %>% 
#   filter(Country %in% focus) %>%
#   group_by(Country) %>%
#   summarise_at(vars(HappinessScore),
#                list(min = min, max=max)) %>%
#   mutate(diff = max-min) %>%
#   arrange(diff) %>%
#     ungroup()

# Set focus countries
focus <- c("Poland", "Germany", "Ukraine", "Romania", "Bulgaria")

# Set up dataframe
data_bubble <- data %>% 
  filter(!is.na(data$HappinessScore) & Year > 2010 & continent == "Europe") %>%
  mutate(mask= case_when(Country %in% focus ~ Country))
  

# produce animation
bubble <- ggplot(data_bubble, aes(gdp_capita, HappinessScore, size = pop, color = mask)) +
  geom_point(data = subset(data_bubble, !is.na(mask)),
             aes(gdp_capita, HappinessScore, size = pop, color = mask),
             alpha = 0.8,show.legend = FALSE) + # Points in Focus
  geom_point(data = subset(data_bubble, is.na(mask)),
             aes(gdp_capita, HappinessScore, size = pop, color = mask),
             alpha = 0.8,show.legend = FALSE) + # Points without Focus
  scale_size(range = c(2, 12)) +
  scale_x_log10() +
  theme(plot.title = element_text(size = 20, hjust = 0.5)) +
  labs(title = 'Happiness vs. GDP per Capita Over Time in {round(frame_time,0)}', 
       x = 'GDP per capita', y = 'Happiness',
       caption = "Own creation. Data: World Happiness Report, World Bank") +
  geom_text(data = subset(data_bubble, Country %in% focus), aes(label = Country), size = 2, color="black") +
  # gganimate specific:
  labs(title = 'Year: {frame_time}', x = 'GDP per Capita', y = 'Happiness') +
  transition_time(as.integer(Year)) +
  ease_aes('linear') +
  shadow_wake(1, alpha = 0.2, exclude_layer = 2:3, size = 1, wrap = FALSE) # no wake for non focus countries and text

# Save as gif (commented to save runtime when knitting)
# gif <- animate(bubble,renderer = gifski_renderer(), end_pause=100, nframes = 200)
# anim_save("img/BubbleChart.gif",animation=gif) 
```

![](img/BubbleChart.gif)

## Happiness and COVID {.flexbox .vcenter}

```{r line plot, echo=FALSE}
# Data prep
data_line <- data %>%
  group_by(Year, continent) %>%
  summarise_at(vars(HappinessScore),
               list(HappinessScore = mean), na.rm = TRUE) %>% # Get mean by continent per year
  ungroup()

data_line2 <- data %>%
  group_by(Year) %>%
  summarise_at(vars(HappinessScore),
               list(HappinessScore = mean), na.rm = TRUE) %>% # Get global mean by year
  ungroup()

data_line2 <- data_line2 %>% add_column("Country"="World") # Add column Country for unioning

names(data_line) <- c("Year",
                      "Country",
                      "HappinessScore") # Unify naming
# Union data
data_line <- union_all(data_line, data_line2)
data_line <- union_all(data_line, filter(data[c("Year", "Country","HappinessScore")], Country == "Poland"))

# Generate Plot
ggplot(filter(data_line, Country %in% c("World","Europe","Poland") & Year >= 2010),aes(Year,HappinessScore, group = Country, color = Country)) +
  geom_rect(aes(xmin = 2019, ymin = -Inf,
                xmax = Inf, ymax = Inf),
            alpha = .4, fill = "grey", color = NA # Highlight COVID time with rectangle
  ) +
  geom_line(size = 1.2) +
  geom_point() +
  geom_vline(xintercept = 2019,color = "black", size = 2) + # Add vertical line to separate pre and post COVID
  geom_dl(aes(label = Country), method = list(dl.combine("last.points"), vjust = 1, hjust=0.5, cex = 1)) +
  annotate("text", x = 2019, y = 7, label = "COVID-19", vjust = 2, hjust=-0.1, size = 5) + # Add text for  COVID
  theme(legend.position = 'bottom',
        plot.title = element_text(size = 20, hjust = 0.5)) +
  labs(color = NULL,
       title = "Happiness Under COVID-19",
       caption = "Own creation. Data: World Happiness Report, World Bank") +
  scale_color_manual(values = c('#003399', '#D22630', 'black')) # Change colours to European blue and Polish red
```

## Fin.{.flexbox .vcenter}
![](img/Happy.png)

## Appendix

## Appendix 1 - Happiness Distribution by Continent {.flexbox .vcenter}

```{r Boxplot, echo=FALSE}
ggplot(data = filter(data, !is.na(data$continent)),
       aes(y = HappinessScore, x = continent)) +
  geom_boxplot() +
  stat_summary(geom = 'point', shape = 15, fun = mean, size = 2) + 
  labs(title = "Happiness Compared by Continent",
       x = "",
       caption = "Own creation. Data: World Happiness Report, World Bank") + 
  theme(legend.position = 'bottom',
        plot.title = element_text(size = 20, hjust = 0.5)) 
```