table(sales_clean$Gender)

gender_freq <- as.data.frame(table(sales_clean$Gender))
gender_freq <- gender_freq[gender_freq$Var1 %in% c("Laki-laki","Perempuan"), ]
gender_freq

ggplot(data = gender_freq, mapping = aes(x = Var1, y = Freq)) +
  geom_col(fill = "pink") +
  geom_label(aes(label = Freq), col = "black") +
  labs(title = "Demografi Usia pelanggan Susu", subtitle = "Pada Channel E-Commerce di Indonesia", x = "Gender", y = "Total") +
  theme_minimal()

gender_freq2 <- as.data.frame(table(sales_clean$Gender, sales_clean$Brand))
gender_freq2 <- gender_freq2[gender_freq2$Var1 %in% c("Laki-laki","Perempuan"), ]
gender_freq2

ggplot(data = gender_freq2, mapping = aes(x = Var1, y = Freq, fill = Var2)) +
  geom_col()+
  geom_label(aes(label = Freq), col = "black") +
  labs(title = "Demografi Usia pelanggan Susu", subtitle = "Pada Channel E-Commerce di Indonesia", x = "Gender", y = "Total") +
  facet_wrap(~ Var2, ncol = 2) +
  scale_fill_manual(values=c('tomato', 'red')) +
  theme(legend.position = 'none')

age_freq <- as.data.frame(table(sales_clean$age_cat))
age_freq

ggplot(data = age_freq, mapping = aes(x = Var1, y = Freq)) +
  geom_col(fill = "tomato") +
  labs(title = "Age Customer Distribution",
       subtitle = "From e-commerce Channel",
       x = "Rentang usia",
       y = NULL) +
  theme_minimal() +
  theme(plot.title = element_text(colour = "darkgreen", hjust = 0.5, face = "bold", size = 20),
        plot.subtitle = element_text(hjust = 0.5))

district_freq <- as.data.frame(table(sales_clean$District))

# ordering
district_freq <- district_freq[order(district_freq$Freq, decreasing = T), ]
names(district_freq) <- c("District", "Freq")
district_freq

ggplot(data = district_freq[1:10, ], 
       aes(x = Freq, y = reorder(District, Freq))) +
  geom_col(fill = "darkseagreen")

ggplot(data = district_freq[1:10, ], 
       aes(x = Freq, y = reorder(District, Freq))) +
  geom_col(fill = "darkseagreen") +
  geom_col(fill = "darkgreen", data = district_freq[2,])

ggplot(data = district_freq[1:10, ], 
       aes(x = Freq, y = reorder(District, Freq))) +
  geom_col(fill = "darkseagreen") +
  geom_col(fill = "darkgreen", data = district_freq[2,]) +
  geom_label(data = district_freq[1:3,], 
             aes(label = Freq), nudge_x = 2)

ggplot(data = district_freq[1:10, ], 
       aes(x = Freq, y = reorder(District, Freq))) +
  geom_col(fill = "darkseagreen") +
  geom_col(fill = "darkgreen", data = district_freq[2,]) +
  geom_label(data = district_freq[1:3,], 
             aes(label = Freq), nudge_x = 2) +
  geom_vline(xintercept = mean(district_freq$Freq),
             linetype = 2,
             lwd = 1,
             color = "tomato") + # memberikan garis mean
  geom_label(aes(label = round(mean(district_freq$Freq),2),
                 x = 19,
                 y = "DUREN SAWIT"),
             fill = "tomato") # memberikan label mean

ggplot(data = district_freq[1:10, ], 
       aes(x = Freq, y = reorder(District, Freq))) +
  geom_col(fill = "darkseagreen") +
  geom_col(fill = "darkgreen", data = district_freq[2,]) +
  geom_label(data = district_freq[1:3,], 
             aes(label = Freq), nudge_x = 2) +
  geom_vline(xintercept = mean(district_freq$Freq),
             linetype = 2,
             lwd = 1,
             color = "tomato") + # memberikan garis mean
  geom_label(aes(label = round(mean(district_freq$Freq),2),
                 x = 19,
                 y = "DUREN SAWIT"),
             fill = "tomato") + # memberikan label mean
  scale_x_continuous(breaks = seq(0, 140, 20))

ggplot(data = district_freq[1:10, ], 
       aes(x = Freq, y = reorder(District, Freq))) +
  geom_col(fill = "darkseagreen") +
  geom_col(fill = "darkgreen", data = district_freq[2,]) +
  geom_label(data = district_freq[1:3,], 
             aes(label = Freq), nudge_x = 2) +
  geom_vline(xintercept = mean(district_freq$Freq),
             linetype = 2,
             lwd = 1,
             color = "tomato") + # memberikan garis mean
  geom_label(aes(label = round(mean(district_freq$Freq),2),
                 x = 19,
                 y = "DUREN SAWIT"),
             fill = "tomato")+ # memberikan label mean
  scale_x_continuous(breaks = seq(0, 140, 20)) +
  labs(title = "Top 10 District",
       subtitle = "Based on Customer Frequency",
       x = "Frequency",
       y = NULL)

ggplot(data = district_freq[1:10, ], 
       aes(x = Freq, y = reorder(District, Freq))) +
  geom_col(fill = "darkseagreen") +
  geom_col(fill = "darkgreen", data = district_freq[2,]) +
  geom_label(data = district_freq[1:3,], 
             aes(label = Freq), nudge_x = 2) +
  geom_vline(aes(xintercept = mean(district_freq$Freq)),
             linetype = 2,
             lwd = 1,
             color = "tomato") + # memberikan garis mean
  geom_label(aes(label = round(mean(district_freq$Freq),2),
                 x = 19,
                 y = "DUREN SAWIT"),
             fill = "tomato")+ # memberikan label mean
  scale_x_continuous(breaks = seq(0, 140, 20)) +
  labs(title = "Top 10 District",
       subtitle = "Based on Customer Frequency",
       x = "Frequency",
       y = NULL) +
  theme_minimal() +
  theme(axis.text.y = element_text(face = "bold", color = "dimgrey"),
        plot.title = element_text(face = "bold", size = 15))

province_freq <- as.data.frame(table(sales_clean$Province))

# Diurutkan
province_freq <- province_freq[order(province_freq$Freq, decreasing = T), ]

# Melakukan pengubahan nama kolom
names(province_freq) <- c("Province", "Freq")

province_freq

ggplot(province_freq, aes(x = reorder(Province, -Freq), y = Freq)) +
  geom_bar(stat = "identity", aes(fill = Province), show.legend = F, width = 0.6) +
  scale_fill_brewer(palette = 5) +
  geom_text(aes(label = Freq), nudge_y = 50) +
  labs(title = "Province Customer Distribution",
       subtitle = "from e-commerce channel",
       y = NULL,
       x = "Province") +
  theme_minimal() +
  theme(plot.title = element_text(face = "bold")
        )

province_agg <- sales_clean[sales_clean$AmountwithPPn <= 2000000 & sales_clean$AmountwithPPn >= 1000000,]
head(province_agg)

ggplot(data = province_agg, aes(x = Province, y = AmountwithPPn)) +
  geom_boxplot(outlier.color = "red") +
  geom_jitter(aes(size = Qty))

ggplot(data = province_agg, aes(x = Province, y = AmountwithPPn)) +
  geom_boxplot(outlier.shape = NA, fill = "khaki", color = "darkgrey") +
  geom_jitter(aes(size = Qty, color = Province)) +
  scale_color_brewer(palette = 8) +
  theme_minimal()

ggplot(data = province_agg, aes(x = Province, y = AmountwithPPn)) +
  geom_boxplot(outlier.shape = NA, fill = "khaki", color = "darkgrey") +
  geom_jitter(aes(size = Qty, color = Province)) +
  scale_color_brewer(palette = 8) +
  labs(title = "Boxplot Distribution of Province",
       subtitle = "Based on Amount With PPN & Quantity",
       x = NULL,
       y = "Amount with PPN (in Rupiah)",
       size = "Quantity") +
  theme_minimal()

library(scales)
ggplot(data = province_agg, aes(x = Province, y = AmountwithPPn)) +
  geom_boxplot(outlier.shape = NA, fill = "khaki", color = "darkgrey") +
  geom_jitter(aes(size = Qty, color = Province)) +
  scale_color_brewer(palette = 8) +
  scale_y_continuous(labels = comma) +
  labs(title = "Boxplot Distribution of Province",
       subtitle = "Based on Amount With PPN & Quantity",
       x = NULL,
       y = "Amount with PPN (in Rupiah)",
       size = "Quantity") +
  theme_minimal()

source_freq <- as.data.frame(table(sales_clean$Source))
source_freq <- source_freq[order(source_freq$Freq, decreasing = T), ]
source_freq <- head(source_freq, 5)
names(source_freq) <- c("Source", "Freq")
source_freq

ggplot(data = source_freq, mapping = aes(x = Freq, y = reorder(Source, Freq))) +
  geom_col(aes(fill = Freq), show.legend = F) +
  scale_fill_gradient(low = "plum", high = "plum4") +
  geom_text(aes(label = Freq), nudge_x = 80) +
  labs(title = "Customer by Source",
       subtitle = "from e-commerce channel",
       x = "Frequency",
       y = NULL) +
  theme_minimal() +
  theme(plot.title = element_text(face = "bold", hjust = 0.4),
        plot.subtitle = element_text(hjust = 0.4))

agg_sales <- aggregate(AmountwithPPn ~ year_purchased + month_purchased,
                       data = sales_clean,
                       FUN = median)
agg_sales <- agg_sales[agg_sales$year_purchased %in% c(2020, 2021), ]
agg_sales$year_purchased <- as.factor(agg_sales$year_purchased)

ggplot(data = agg_sales, aes(x = month_purchased, y = AmountwithPPn)) +
  geom_line(aes(group = year_purchased, color = year_purchased), size = 1) +
  geom_point(aes(color = year_purchased), size = 2) +
  scale_x_discrete(guide = guide_axis(angle = 45)) +
  scale_color_brewer(palette = "Set2")

ggplot(data = agg_sales, aes(x = month_purchased, y = AmountwithPPn)) +
  geom_line(aes(group = year_purchased, color = year_purchased), size = 1) +
  geom_point(aes(color = year_purchased), size = 2) +
  geom_label(data = agg_sales[agg_sales$month_purchased == 'July', ], 
             aes(label = scales::comma(round(AmountwithPPn, 1))), 
             size = 2, 
             nudge_y = 10000) +
  scale_x_discrete(guide = guide_axis(angle = 45)) +
  scale_y_continuous(labels = scales::comma) +
  scale_color_brewer(palette = "Set2")

ggplot(data = agg_sales, aes(x = month_purchased, y = AmountwithPPn)) +
  geom_line(aes(group = year_purchased, color = year_purchased), size = 1) +
  geom_point(aes(color = year_purchased), size = 2) +
  geom_label(data = agg_sales[agg_sales$month_purchased == 'July', ], 
             aes(label = scales::comma(round(AmountwithPPn, 1))), 
             size = 2, 
             nudge_y = 10000) +
  scale_x_discrete(guide = guide_axis(angle = 45)) +
  scale_y_continuous(labels = scales::comma) +
  scale_color_brewer(palette = "Set2") +
  labs(title = "Monthly Trend Sales Revenue",
       subtitle = "in 2020 and 2021",
       y = "Month",
       x = "Amount of Sales", 
       color = "Year") +
  theme_minimal() +
  theme(plot.title = element_text(hjust = 0.5, face = "bold"),
        plot.subtitle = element_text(hjust = 0.5))

source_agg <- aggregate(AmountwithPPn ~ Source,
                        data = sales_clean,
                        FUN = sum)
source_agg <- source_agg[order(source_agg$AmountwithPPn, decreasing = T), ]
source_agg$percentage <- source_agg$AmountwithPPn/sum(source_agg$AmountwithPPn)
source_agg

ggplot(data = source_agg, aes(x = percentage, y = reorder(Source, percentage))) +
  geom_col(aes(fill = percentage), show.legend = F) +
  geom_label(aes(label = scales::percent(round(percentage,4))), size = 3, nudge_x = 0.05) +
  scale_fill_gradient(low = "peru", high = "moccasin") +
  scale_x_continuous(label = scales::percent) +
  labs(title = "Contribution Sales by Source",
       subtitle = "in percentage (%)",
       y = "Source") +
  theme_minimal() +
  theme(plot.title = element_text(hjust = 0.35, face = "bold"),
        plot.subtitle = element_text(hjust = 0.35))

product_agg <- aggregate(AmountwithPPn ~ Product + Brand,
                        data = sales_clean,
                        FUN = sum)
product_agg <- product_agg[order(product_agg$AmountwithPPn, decreasing = T), ]
product_agg

library(scales)
ggplot(data = product_agg, mapping = aes(y = reorder(Product,AmountwithPPn), x = AmountwithPPn)) +
  geom_col(aes(fill = Brand), show.legend = F) +
  geom_text(aes(label = comma(AmountwithPPn)), size = 2) +
  facet_wrap(~Brand, ncol = 2, scales = "free") +
  # untuk mengatur wraping nilai agar tidak terlalu panjang
  scale_y_discrete(label = label_wrap(20)) +
  # untuk mengatur agar skala x ada koma
  scale_x_continuous(label = comma) +
  # untuk mengatur ukuran axis x
  theme(axis.text.y = element_text(size = 7),
        axis.text.x = element_text(size = 5)) +
  labs(title = "Contribution Sales by Product",
       subtitle = "in Total",
       y = "Product Name",
       x = NULL)

# melakukan filtering hanya untuk Province DKI Jakarta
city_jakarta <- sales_clean[sales_clean$Province == "DKI JAKARTA", ]

# Melakukan Aggregate dan pembentukan percent
city_agg1 <- aggregate(AmountwithPPn ~ City + Brand,
          data = city_jakarta,
          FUN = mean)
city_agg1 <- transform(city_agg1,
                       perc = ave(AmountwithPPn,
                                  City,
                                  FUN = prop.table))
city_agg1

ggplot(data = city_agg1, mapping = aes(x = AmountwithPPn, y = City))+
  geom_col(aes(group = Brand, fill = Brand), position = position_fill(reverse = T)) +
  scale_x_continuous(label = percent) +
  geom_text(aes(label = percent(perc)), position = position_fill(vjust = 0.5), size = 3.1) +
  scale_fill_brewer(palette = "Set3") +
  theme_minimal()

# Melakukan Aggregate dan pembentukan percent
prov_agg1 <- aggregate(AmountwithPPn ~ Province + Brand,
          data = sales_clean,
          FUN = mean)
prov_agg1

ggplot(data = prov_agg1, aes(x = Province, y = AmountwithPPn, fill = Brand)) +
  geom_col(aes(group = Brand), position = "dodge") +
  geom_text(aes(label = comma(x = AmountwithPPn, prefix = "Rp.", suffix = ",-")), 
            position = position_dodge(0.9), vjust = -0.5, size = 2.9) +
  scale_y_continuous(labels = comma, breaks = seq(0,600000,100000), limits = c(0, 500000)) +
  theme_minimal()

library(leaflet)

city_all <- aggregate(AmountwithPPn ~ City,
          data = sales_clean,
          FUN = mean)
city_all$percent <- city_all$AmountwithPPn/sum(city_all$AmountwithPPn)
city_all

# menambahkan data latitude_longitude
latlong <- data.frame("lat" = c(-6.9147444, -6.8937121, -6.233333, -6.6, -7.3333333, -6.8172531, -6.0169825, -6.880239, -6.715534, -6.39, -7.227906,    -6.336315, -6.211544, -6.211544,    -6.332973, -6.211544, -6.211544, -6.3227303, -6.9833333, -6.8531026, -6.314835, -6.5386806, -6.12009, -6.569361, -6.92405,  -6.8329, -6.1783056, -6.2888889, -7.327954), 
                         
                         "long" = c(107.6098111, 107.4321959, 107, 106.8, 108.35, 107.1307289, 106.040506, 107.5355, 108.564003, 106.83, 107.908699, 108.325104, 106.845172, 106.845172, 106.807915, 106.845172, 106.845172, 107.3375791, 108.4833333, 108.2258897, 106.103897, 107.4499404, 106.150299,    107.752403, 106.922203, 107.9532, 106.6318889,  106.7180556, 108.214104))

# menggabungkan dua kolom data
city_all_clean <- cbind(city_all, latlong)

library(maps)
global <- map_data("world")
world <- ggplot() + 
  geom_polygon(data = global, aes(x=long, y = lat, group = group), 
               fill = "gray85", color = "gray80") +
  coord_fixed(1.3)

Indo <- world + xlim(94,142) + ylim(-11,7.5)

library(ggrepel)
world + xlim(105,115) + ylim(-9,-5.8)+ 
  geom_point(data = city_all_clean, aes(x = long, y = lat, size = AmountwithPPn, color = AmountwithPPn),
             alpha = 0.5, 
             show.legend = F) +
  scale_color_gradient(low = "yellow", high = "red") +
  geom_text_repel(data = city_all_clean, aes(x = long, y = lat, label= City), size = 2)

library(htmltools)
library(dplyr)
library(leaflet)
m <- leaflet(city_all_clean)


pop <- paste("Region: ", city_all_clean$City, "<br/>",
             "Revenue: ", round(city_all_clean$AmountwithPPn)) %>% 
  lapply(htmltools::HTML)

# mendefinisikan variabel yang menjadi dasar pewarnaan
col <- city_all_clean$AmountwithPPn

mybins <- seq(50000, 700000, by=150000)
pal <- colorBin( palette="YlOrRd", 
                       domain=col, 
                       bins=mybins)

m %>% 
  addProviderTiles(providers$CartoDB.DarkMatter) %>% 
  addCircleMarkers(~long, ~lat, 
                   fillColor = ~pal(col),
                   weight = 1,
                   color = "white",
                   fillOpacity = 0.8,
                   radius=8, 
                   stroke=FALSE,
                   label = pop,) %>%
  addLegend("bottomright", 
            pal = pal,
            values = ~col,
            title = "Revenue",
            labFormat = labelFormat(digits = 2),
            opacity = 1)