plot_col_top10listnb <- df_list_summary %>%
    group_by(neighbourhood) %>%
    summarize(num_listings = n(), neighbourhood_grp = unique(neighbourhood_group)) %>%
    arrange(desc(num_listings)) %>% 
  mutate(Text=paste(
         "Region: ",neighbourhood_grp, "<br>",
         "Planning Area: ",neighbourhood, "<br>",
         "No. of Listings: ",num_listings, sep = ""
      )) %>% 
    head(10) %>% 
  
ggplot(aes(x = reorder(neighbourhood, num_listings), 
           y = num_listings, 
           text = Text, fill = neighbourhood_grp)) +
    geom_col() +
    coord_flip() +
    theme(legend.position = "bottom") +
    labs(x = "", y = "Homestay Count",
         fill = "Region") +
    theme(
      panel.border = element_blank(),
      panel.grid.major = element_blank()
    )

ggplotly(plot_col_top10listnb, tooltip="text") %>%
  layout(title = list(text = paste0('Top 10 Planning Area with Highest Number of Airbnb Homestays',
                                    '<br>',
                                    '<sup>',
                                     'among all Region in Singapore','</sup>')),
         legend = list(orientation = "h", x = 0.1, y = -0.3)) %>%
   config(displayModeBar = F)

agg_homestayperRegion <- df_list_summary %>% 
  group_by(neighbourhood) %>% 
  summarize(neighbourhood_group = unique(neighbourhood_group), num_listings = n()) %>% 
  arrange(desc(neighbourhood_group), desc(num_listings))


agg_homestayperRegion %>% group_by(neighbourhood_group) %>% 
  slice_max(num_listings, n = 3, with_ties = F) %>%
  arrange(desc(neighbourhood_group)) %>% 
  ggplot(aes(x = num_listings, y = reorder(neighbourhood, num_listings))) +
  geom_col(aes(fill = neighbourhood), show.legend = F) +
  geom_label(mapping = aes(label = num_listings),
             position = position_dodge(width = 0.5), 
             #nudge_y = 3, 
             size = 2.5, 
             show.legend = F, fill = "white") +
  facet_wrap(facets = ~neighbourhood_group, nrow = 5, scales = "free_y") +
  theme_linedraw() + scale_fill_discrete() +
  labs(title = "Top 3 Planning Area with Highest Number of Homestay",
       subtitle = "per Region in Singapore",
       x = "Homestay Count",
       y = "Planning Area",
       fill = "Frequency") +
  theme(panel.grid.major.y = element_blank(), 
        panel.grid.minor.y = element_blank(), 
        panel.grid.minor.x = element_blank(),
        panel.grid.major.x = element_line(colour = "grey", linetype = "dashed"))

homestaydist_perRegion <- df_list_summary %>% 
  group_by(neighbourhood_group) %>% summarize(num_listings = n()) %>%
  mutate(percentage = round((num_listings/sum(num_listings))*100,2))


ggplot(homestaydist_perRegion, 
       aes(x=num_listings,
           y=reorder(neighbourhood_group, num_listings))) +
  geom_col(aes(fill=neighbourhood_group, show.legends = FALSE)) +
  coord_flip() +
  geom_hline(yintercept = mean(homestaydist_perRegion$num_listings), linetype = 3) +
  geom_label(aes(label = paste0(num_listings, ' ','(',percentage,' %',')'
                                )),
             size = 3, 
             show.legend = F, fill = "white") + theme_bw() +
  labs(title = "Top 10 Region in Singapore",
       subtitle = "based on number of homestay",
       x = "Number of Homestay", y = "Region") +
  theme(
    panel.border = element_blank(),
    panel.grid.minor = element_blank(),
    legend.position = "none"
  )

# Plot host dist in map
temp_height <- max(df_list_summary$latitude) - min(df_list_summary$latitude)
temp_width <- max(df_list_summary$longitude) - min(df_list_summary$longitude)
temp_borders <- c(bottom  = min(df_list_summary$latitude)  - 0.05 * temp_height,
                  top     = max(df_list_summary$latitude)  + 0.05 * temp_height,
                  left    = min(df_list_summary$longitude) - 0.1 * temp_width,
                  right   = max(df_list_summary$longitude) + 0.1 * temp_width)

#load staten map
temp_map <- get_stamenmap(temp_borders, zoom = 12, maptype = "toner-lite")

#load ggmap + aes geom point
plot_hostpointmap <- ggmap(temp_map) +
  geom_point(data = df_list_summary, 
             mapping = aes(x = longitude, y = latitude,
            col = neighbourhood_group, alpha=0.7)) +
  labs(x = "longitude",
       y = "latitude", 
       title = "Homestay Distribution in Singapore",
       col = "Region") +
  guides(alpha = FALSE) +
  theme(plot.title = element_text(vjust=5))

plot_hostpointmap

kable(unique(df_list_summary$room_type))

df_list_summary %>%
  count(room_type, sort = TRUE) %>%
  mutate(room_type = reorder(room_type, n), percentage = round((n/sum(n))*100,2)) %>%
  
  ggplot(aes(room_type, n)) +
    geom_col(aes(fill = room_type)) +
  coord_flip() +
  geom_label(aes(label = paste0(n, ' ', '(',percentage,' %',')')),
             nudge_y = 1, size = 2.5, show.legend = F, fill = "white") +
  labs(title = "Aibnb Type of Place Distribution",
       subtitle = "over Singapore",
       x=NULL, y= "Count", fill = "Type of Place") + 
  theme_bw() +
  theme(
    panel.border = element_blank(),
    panel.grid.minor = element_blank(),
    legend.position = "bottom"
  )

plot_agg_roomtypeperRegion <- data.frame(table(region = df_list_summary$neighbourhood_group,
                                               room_type = df_list_summary$room_type)) %>% 
mutate(Text=paste(
         "Region: ",region, "<br>",
         "Type of Place: ",room_type, "<br>",
         "Count:",Freq, sep = ""
      )) %>%
  
ggplot(aes(x=region, y=Freq, group=room_type, text=Text)) +
      geom_col(aes(fill=room_type), position = "dodge") +
      labs(x = "",
           y = "Count",
           fill = "Type of Place") +
      theme_bw()+
      theme(
        panel.border = element_blank()
      )
  
ggplotly(plot_agg_roomtypeperRegion, tooltip="text") %>%
  layout(legend = list(orientation = "h", x = 0.1, y = -0.05),
         title = list(text = paste0('Type of Place Distribution',
                                    '<br>',
                                    '<sup>','per Region','</sup>')))

plot_pricedist <- ggplot(df_list_summary, aes(price)) +
  # using bins = 50 and continous scale
  geom_histogram(aes(y = ..density..), col = "black", fill = "white") + 
  geom_density(alpha = 0.3, fill="#FF6666") +
  geom_vline(xintercept = round(mean(df_list_summary$price), 2),linetype = 3) +
  scale_x_log10() +
  labs(title =  "Transformed distribution of Price",
      subtitle = expression("With" ~'log'[10] ~ "transformation of x-axis"),
      x="Price", y = "Density", 
      fill = "Type of Place") +
 guides(col = F)

library(viridis)
plot_price_per_roomtype <- df_list_summary %>% select(price, room_type) %>% 
  ggplot(aes(price, fill = room_type))+
  geom_density(aes(col = room_type, show.legends = F), alpha = 0.5, round=2) +
  scale_fill_viridis(discrete = TRUE) +
  scale_color_viridis(discrete = TRUE) +
  theme(text = element_text(size = 10)) +
  scale_x_continuous(limits = c(0, 1000), breaks = seq(0, 1000, 100))+
  labs(title = "Price Distribution",
       subtitle = "per Type of Place",
       x="Price", y = "Density", 
       fill = "") +
  guides(col = F) + theme(legend.position = "bottom") +
  labs(title = "S'pore Airbnb Price Distribution",
       subtitle = "by Type of Place",
       x="Price", y= "Density")

subplot1 <- ggarrange(plot_pricedist, plot_price_per_roomtype,
                      ncol = 2, nrow = 1, 
                      common.legend = FALSE, 
                      legend = "bottom")
subplot1

#df_list_summary %>% names()
plot_boxplot_prt <- df_list_summary %>%
  ggplot(aes(x = price, y = room_type)) +
      geom_boxplot(aes(fill = room_type)) +
      scale_x_log10() +
      labs(x = expression( ~'log'[10] ~ "Price"), y = "",
           title = "Price Distribution",
           subtitle = "per Type of Place",
           fill = "Type of Place") +
      theme_bw() +
      theme(
        axis.text.x = element_text(angle = 45, hjust = 1)
      )
plot_boxplot_prt

library(RColorBrewer)
# Set the color scale
palette <- brewer.pal(5, "RdYlBu")[-(2:4)]

df_list_summary %>% 
  group_by(room_type) %>% 
  summarise(avg_price = round(mean(price),2)) %>% 
  ggplot(aes(x = avg_price, y = room_type, color = avg_price)) +
  geom_segment(aes(xend = 10, yend = room_type), size = 2) +
  geom_point(aes(color = avg_price),size = 10) +
  geom_text(aes(label = avg_price), color = "white", size = 2.5) +
  scale_x_continuous("", limits = c(10, 250), position = "top") +
  scale_color_gradientn(colors = palette) +
  labs (
    title = "Average Price per Room Type", 
    fill = "Average Price"
  ) +
  theme_classic() +
  theme(axis.line.y = element_blank(),
        axis.ticks.y = element_blank(),
        axis.text = element_text(color = "black"),
        axis.title = element_blank(),
        legend.position = "none")

df_est_revenue <- df_list_summary %>% 
  select(-c(latitude, longitude, last_review)) %>% 
  mutate(est_revenue = minimum_nights * price * number_of_reviews) %>%
  arrange(desc(est_revenue))

kable(head(df_est_revenue,10))

agg_est_rev <- df_est_revenue %>% 
  group_by(room_type) %>% 
  summarise(avg.price = round(mean(price),2),
            avg.min.nights = round(mean(minimum_nights),2),
            avg.nbr.reviews = round(mean(number_of_reviews),2),
            avg.avail.365 = round(mean(availability_365),2),
            avg.estimated.revenue = round(mean(est_revenue),2)
            )
kable(agg_est_rev)

plot_est1 <- agg_est_rev %>% 
  ggplot(aes(x = avg.estimated.revenue, y = reorder(room_type, avg.estimated.revenue), color = avg.estimated.revenue)) +
  geom_segment(aes(xend = 10, yend = room_type), size = 15) +
  geom_text(aes(label = avg.estimated.revenue), color = "black", size = 3, 
            nudge_x = 4000) +
  scale_x_continuous(name = "", 
                     labels = scales::comma, position = "top") +
  scale_color_gradientn(colors = palette) +
  labs (
    title = "Average Estimated Revenue (in USD)", 
    subtitle = "by Type of Place",
    fill = ""
  ) +
  theme_classic() +
  theme(axis.line.y = element_blank(),
        axis.ticks.y = element_blank(),
        axis.text = element_text(color = "black"),
        axis.title = element_blank(),
        legend.position = "none")

plot_est2 <- agg_est_rev %>% 
  ggplot(aes(x = avg.nbr.reviews, y = reorder(room_type, avg.estimated.revenue), )) +
  geom_col(aes(fill = avg.nbr.reviews)) + 
  geom_text(aes(label = avg.nbr.reviews), color = "white", size = 3, 
            nudge_x = -2) +
  labs (
    title = "Avg. Number of Reviews", 
    subtitle = "by Type of Place",
    fill = "") +
  theme_classic() +
  theme(axis.line.y = element_blank(),
        axis.ticks.y = element_blank(),
        axis.text = element_text(color = "black"),
        axis.title = element_blank(),
        legend.position = "none")

plot_est3 <- agg_est_rev %>% 
  ggplot(aes(x = avg.min.nights, y = reorder(room_type, avg.estimated.revenue))) +
  geom_col(aes(fill = avg.min.nights)) + 
  geom_text(aes(label = avg.min.nights), color = "white", size = 2.5, 
            nudge_x = -0.9) +
  labs (
    title = "Avg. Minimum Number of \nNights Spend", 
    subtitle = "by Type of Place",
    fill = "", y = "qwe", x = "") +
  theme_classic() +
  theme(axis.line.y = element_blank(),
        axis.ticks.y = element_blank(),
        axis.text = element_text(color = "black"),
        axis.title = element_blank(),
        legend.position = "none")

grid.arrange(plot_est1, plot_est2, plot_est3,
             nrow=2,
             layout_matrix = rbind(c(1,1),c(2,3)))

plot_PricevsReviews <- df_list_summary %>% 
  select(c(id, host_id, number_of_reviews, neighbourhood, neighbourhood_group, 
           reviews_per_month, room_type, price, minimum_nights, availability_365)) %>% 
  group_by(neighbourhood_group) %>% 
  arrange(desc(number_of_reviews)) %>% 
  
  ggplot(aes(x = price, y = number_of_reviews)) +
  geom_point(aes(col = room_type), show.legend = FALSE, alpha = 0.2) +
  theme_bw() +
  labs(x = "price", y = "number of reviews", 
       title = "Price vs Number of Reviews")
  
ggplotly(plot_PricevsReviews) %>%
 layout(title = list(text = paste0('Price vs Number of Reviews')),
        xaxis = list(text = 'number of reviews'), 
        legend = list(title=list(text='<b> Type of Place </b>')))