visualization.R

# Install required packages if not already installed
packages <- c("tidyverse", "ggplot2", "scales")
installed <- packages %in% rownames(installed.packages())
if (any(!installed)) install.packages(packages[!installed])

# Load libraries
library(tidyverse)
library(ggplot2)
library(scales)

# Disable scientific notation
options(scipen = 999)

# Import Cyclistic clean data
cyclistic_df <- read_csv("C:/Users/user/Desktop/Cyclistic_Case_Study/data/csv_data_clean/Cyclistic_Data_Clean_Jul24_to_Jun25.csv")

# Define the order of months from July to June
month_order <- c("Jul", "Aug", "Sep", "Oct", "Nov", "Dec",
                 "Jan", "Feb", "Mar", "Apr", "May", "Jun")

# Convert month_of_year to ordered factor
cyclistic_df <- cyclistic_df %>%
  mutate(month_of_year = factor(month_of_year,
                                levels = month_order,
                                ordered = TRUE))

# Define the order of days from Monday to Sunday
day_order <- c("Mon", "Tue", "Wed", "Thu",
               "Fri", "Sat", "Sun")

# Convert day_of_week to ordered factor
cyclistic_df <- cyclistic_df %>%
  mutate(day_of_week = factor(day_of_week,
                              levels = day_order,
                              ordered = TRUE))

# Creating an ordered factor for the seasons in chronological order
season_order <- c("Spring", "Summer",
                  "Fall", "Winter")

# Convert year_seasons to ordered factor
cyclistic_df <- cyclistic_df %>%
  mutate(year_seasons = factor(year_seasons,
                               levels = season_order,
                               ordered = TRUE))



# PART A ??? Identify GENERAL TRENDS
# Change path directory
setwd("C:/Users/user/Desktop/Cyclistic_Case_Study/Pictures")

# Visualization 1 - User type distribution
cyclistic_df %>%
  ggplot(aes(x = member_casual, fill = member_casual)) +
  geom_bar(position = "dodge") +
  labs(fill = "User Type",
       title = "Total per user type : casual vs member",
       x = "Users",
       y = "Total Users") +
  theme_minimal() +
  theme(plot.title = element_text(hjust = 0.5)) +
  scale_y_continuous(labels = label_comma())

# Save image
ggsave("1_user_type_distribution_chart.png",
       width = 6,
       height = 4,
       dpi = 300)


# Visualization 2 - User type distribution
# Prepare data with percentages
df_pieer <- cyclistic_df %>%
  count(member_casual) %>%
  mutate(
    pourcentage = round(n / sum(n) * 100, 1),
    label = paste0(member_casual, "\n", pourcentage, "%")
  )

df_pieer %>%
  ggplot(aes(x = "", y = n, fill = member_casual)) +
  geom_col(width = 1, color = "white") +
  coord_polar(theta = "y") +
  geom_text(aes(label = label),
            position = position_stack(vjust = 0.5),
            color = "white", size = 4.5) +
  labs(title = "Total per user type : casual vs member",
       fill = "") +
  theme_void() +
  theme(
    plot.title = element_text(hjust = 0.5)
  )


# Save image
ggsave("2_user_type_distribution_pie_chart.png",
       width = 6,
       height = 4,
       dpi = 300)


# Visualization 3 - Total number of rides per month (overall volume)
cyclistic_df %>%
  ggplot(aes(x = month_of_year)) +
  geom_bar(fill = "steelblue") +
  labs(title = "Monthly ride trend",
       subtitle = "July 2024 to June 2025",
       x = "Month",
       y = "Number of rides") +
  theme_minimal() +
  theme(
    plot.title = element_text(hjust = 0.5),
    plot.subtitle = element_text(hjust = 0.5)
  ) + scale_y_continuous(labels = label_comma())

# Save image
ggsave("3_monthly_rides_all_chart.png",
       width = 6,
       height = 4,
       dpi = 300)


# Visualization 4 - Number of rides per season
cyclistic_df %>%
  ggplot(aes(x = year_seasons, fill = year_seasons)) +
  labs(fill = "Seasons year",
       x = "Seasons",
       y = "Number of Rides",
       title = "Number of Rides per season") +
  geom_bar(position = "dodge") +
  theme_minimal() +
  theme(plot.title = element_text(hjust = 0.5)) +
  scale_y_continuous(labels = label_comma())

# Save image
ggsave("4_seasonal_rides_all_chart.png",
       width = 6,
       height = 4,
       dpi = 300)


# Visualization 5 - Busiest days of the week
cyclistic_df %>%
  ggplot(aes(x = day_of_week, fill = day_of_week)) +
  geom_bar(width = 0.6) +
  labs(fill = "Days",
       title = "Number of rides per day of the week",
       x = "Day",
       y = "Number of rides") +
  theme_minimal() +
  theme(plot.title = element_text(hjust = 0.5)) +
  scale_y_continuous(labels = label_comma())

# Save image
ggsave("5_weekday_rides_all_chart.png",
       width = 6,
       height = 4,
       dpi = 300)


# Visualization 6 - Number of rides Weekday vs Weekend
cyclistic_df %>%
  ggplot(aes(x = weekend_vs_weekday, fill = weekend_vs_weekday)) +
  geom_bar(width = 0.6) +
  labs(fill = "Weekday vs Weekend",
       title = "Rides during weekday vs weekend",
       x = "",
       y = "Number of rides") +
  theme_minimal() +
  theme(
    plot.title = element_text(hjust = 0.5),
    axis.text.x = element_text(hjust = 0.5)
  ) +
  scale_y_continuous(labels = label_comma())

# Save image
ggsave("6_rides_weekend_vs_weekday_chart.png",
       width = 6,
       height = 4,
       dpi = 300)


# Visualization 7 - Create pie chart
# Prepare data with percentages
df_pie <- cyclistic_df %>%
  count(weekend_vs_weekday) %>%
  mutate(
    pourcentage = round(n / sum(n) * 100, 1),
    label = paste0(weekend_vs_weekday, "\n", pourcentage, "%")
  )

df_pie %>%
  ggplot(aes(x = "", y = n, fill = weekend_vs_weekday)) +
  geom_col(width = 1, color = "white") +
  coord_polar(theta = "y") +
  geom_text(aes(label = label),
            position = position_stack(vjust = 0.5),
            color = "white", size = 4.5) +
  labs(title = "Ride distribution: weekday vs weekend",
       fill = "") +
  theme_void() +
  theme(
    plot.title = element_text(hjust = 0.5)
  )

# Save image
ggsave("7_rides_share_weekend_vs_weekday_pie_chart.png",
       width = 6,
       height = 4,
       dpi = 300)


# Visualization 8 - Number of rides per hour
cyclistic_df %>%
  ggplot(aes(x = hour_of_day, fill = member_casual)) +
  geom_bar(position = "dodge") +
  labs(fill = "User Type",
       x = "Hour",
       y = "Hourly ride number",
       title = "Hourly ride numbers in a week") +
  facet_wrap(~day_of_week) +
  theme_minimal() +
  theme(plot.title = element_text(hjust = 0.5)) +
  scale_y_continuous(labels = label_comma())

# Save image
ggsave("8_hourly_rides_per_day_user_type_chart.png",
       width = 6,
       height = 4,
       dpi = 300)


# Visualization 9 - Most popular hours of the day
cyclistic_df %>%
  ggplot(aes(x = hour_of_day)) +
  geom_density(fill = "steelblue", size = 0.8) +
  labs(
    title = "Traffic by hour of day",
    subtitle = "From 00:00 to 23:00",
    x = "Hours",
    y = "Traffic concentration"
  ) +
  theme_minimal() +
  theme(
    plot.title = element_text(hjust = 0.5),
    plot.subtitle = element_text(hjust = 0.5)
  )

# Save image
ggsave("9_hourly_traffic_density_chart.png",
       width = 6,
       height = 4,
       dpi = 300)


# Visualization 10 - Bike usage frequencies
cyclistic_df %>%
  ggplot(aes(x = rideable_type, fill = rideable_type)) +
  geom_bar(position = "dodge") +
  labs(fill = "Bike Type",
       x = "Bike",
       y = "Number of bikes",
       title = "Bike usage trends") +
  theme_minimal() +
  theme(plot.title = element_text(hjust = 0.5)) +
  scale_y_continuous(labels = label_comma())

# Save image
ggsave("10_bike_type_usage_chart.png",
       width = 6,
       height = 4,
       dpi = 300)



# PART B ??? User Type Analysis
# Visualization 11 - Number of rides per month
cyclistic_df %>%
  ggplot(aes(x = month_of_year, fill = member_casual)) +
  geom_bar(position = "dodge") +
  labs(fill = "User Type",
       x = "Month",
       y = "Total of rides",
       title = "Number of rides per month",
       subtitle = "July 2024 to June 2025") +
  theme_minimal() +
  theme(
    plot.title = element_text(hjust = 0.5),
    plot.subtitle = element_text(hjust = 0.5)) +
  scale_y_continuous(labels = label_comma())

# Save image
ggsave("11_monthly_rides_by_user_type_chart.png",
       width = 6,
       height = 4,
       dpi = 300)


# Visualization 12 - Rides per season and user type
cyclistic_df %>%
  ggplot(aes(x = year_seasons, fill = member_casual)) +
  labs(fill = "User Type",
       x = "Seasons",
       y = "Number of Rides",
       title = "Number of Rides per season",
       subtitle = "Members preference season") +
  geom_bar(position = "dodge") +
  theme_minimal() +
  theme(
    plot.title = element_text(hjust = 0.5),
    plot.subtitle = element_text(hjust = 0.5)) +
  scale_y_continuous(labels = label_comma())

# Save image
ggsave("12_seasonal_rides_by_user_type_chart.png",
       width = 6,
       height = 4,
       dpi = 300)


# Visualization 13 - Weekday usage by user type
cyclistic_df %>%
  ggplot(aes(x = day_of_week, fill = member_casual)) +
  geom_bar(position = "dodge") +
  labs(fill = "User Type",
       title = "Usage per days of week",
       x = "Days",
       y = "Number of rides") +
  theme_minimal() +
  theme(plot.title = element_text(hjust = 0.5)) +
  scale_y_continuous(labels = label_comma())

# Save image
ggsave("13_weekday_rides_by_user_type_chart.png",
       width = 6,
       height = 4,
       dpi = 300)


# Visualization 14 - Rides per user type: Weekday vs Weekend
cyclistic_df %>%
  count(member_casual, weekend_vs_weekday) %>%
  ggplot(aes(x = weekend_vs_weekday, y = n, fill = member_casual)) +
  geom_col(position = "dodge", width = 0.6) +
  labs(fill = "User Type",
       title = "Rides by user type on weekday and weekend",
       x = "Weekday vs weekend",
       y = "Number of rides") +
  theme_minimal() +
  theme(plot.title = element_text(hjust = 0.5)) +
  scale_y_continuous(labels = label_comma())

# Save image
ggsave("14_rides_weekend_vs_weekday_by_user_type_chart.png",
       width = 6,
       height = 4,
       dpi = 300)


# Visualization 15 - Preferred bike type by user
cyclistic_df %>%
  count(member_casual, rideable_type) %>%
  ggplot(aes(x = rideable_type, y = n, fill = member_casual)) +
  geom_col(position = "dodge", width = 0.6) +
  labs(fill = "User Type",
       title = "Bike type used by user type",
       x = "Bike type",
       y = "Number of rides") +
  theme_minimal() +
  theme(plot.title = element_text(hjust = 0.5)) +
  scale_y_continuous(labels = label_comma())

# Save image
ggsave("15_bike_type_by_user_type_chart.png",
       width = 6,
       height = 4,
       dpi = 300)


# Visualization 16 - Usage by hour per user type
cyclistic_df %>%
  count(member_casual, hour_of_day) %>%
  ggplot(aes(x = hour_of_day, y = n, color = member_casual, group = member_casual)) +
  geom_line(size = 1) +
  labs(color = "User Type",
       title = "Hourly ride distribution",
       subtitle = "From 00:00 to 23:00",
       x = "Hour",
       y = "Number of rides") +
  theme_minimal() +
  theme(plot.title = element_text(hjust = 0.5),
        plot.subtitle = element_text(hjust = 0.5)) +
  scale_y_continuous(labels = label_comma())

# Save image
ggsave("16_hourly_rides_by_user_type_line_chart.png",
       width = 6,
       height = 4,
       dpi = 300)


# Visualization 17 - Area chart per hour per user
cyclistic_df %>%
  count(member_casual, hour_of_day) %>%
  ggplot(aes(x = hour_of_day, y = n, fill = member_casual)) +
  geom_area(position = "stack", alpha = 0.6) +
  labs(fill = "User Type",
       title = "Rides per hour by type (stacked area)",
       subtitle = "From 00:00 to 23:00",
       x = "Hour",
       y = "Number of rides") +
  theme_minimal() +
  theme(plot.title = element_text(hjust = 0.5),
        plot.subtitle = element_text(hjust = 0.5)) +
  scale_y_continuous(labels = label_comma())

# Save image
ggsave("17_hourly_rides_by_user_type_area_chart.png",
       width = 6,
       height = 4,
       dpi = 300)


# Visualization 18 - Average ride duration per user type
cyclistic_df %>%
  group_by(member_casual) %>%
  summarise(duree_moyenne = mean(ride_duration, na.rm = TRUE)) %>%
  ggplot(aes(x = member_casual, y = duree_moyenne, fill = member_casual)) +
  geom_col(width = 0.6) +
  labs(fill = "User Type",
       title = "Average ride duration per user type",
       x = "User Type",
       y = "Average duration (min)") +
  theme_minimal() +
  theme(plot.title = element_text(hjust = 0.5))

# Save image
ggsave("18_average_ride_duration_by_user_type_chart.png",
       width = 6,
       height = 4,
       dpi = 300)



# Finding the most popular start stations among all users
View (cyclistic_df %>%
        filter(start_station_name != "unknown_name") %>%
        group_by(start_station_name) %>%
        summarise(count = n()) %>%
        arrange(desc(count)))


# Finding the most popular start stations for casual users only
View (cyclistic_df %>%
        filter(member_casual == 'casual', start_station_name != "unknown_name" ) %>%
        group_by(member_casual, start_station_name) %>%
        summarise(count = n()) %>%
        arrange(desc(count)))


# Finding the most popular start stations for members only
View (cyclistic_df %>%
        filter(member_casual == 'member', start_station_name != "unknown_name") %>%
        group_by(member_casual, start_station_name) %>%
        summarise(count = n()) %>%
        arrange(desc(count)))