Analysis of Bike Sharing Company - Cyclistic.Rmd

---
title: 'Analysis of Bike Sharing Company: Cyclistic'
author: "Wale Adio"
date: "2023-08-19"
output: 
  html_document: 
    toc: yes
    number_sections: yes
---

```{r setup, include=FALSE}
library(tidyverse)
library(lubridate)
library(ggplot2)
library(dplyr)
knitr::opts_chunk$set(echo = TRUE)
# Set the root directory for notebook chunks
knitr::opts_knit$set(root.dir = "/Users/Wale/Downloads/Projects/Google Analytics/Cyclistic/Divvy_Project")
```

# Data Analysis
## Installing and Loading Packages
Analysis commenced by gathering and cleaning data sets from multiple quarters **(2019 q2, q3, q4 & 2020 q1)** to ensure uniform data. The `Tidyverse` package was utilized for data importation and wrangling. The `Lubridate` package was utilized for data handling attributes, while `Dplyr` and `Ggplot2` packages were utilized for data manipulation and visualization respectively.


## Setting Directory
First the directory needs to be established for data importation

```{r}
setwd("/Users/Wale/Downloads/Projects/Google Analytics/Cyclistic/Divvy_Project")

```

## Importing Datasets
The data set were imported, with the data set capturing bike sharing data from 2019 q2 to 2020 q1. 

```{r echo=TRUE, message=FALSE, warning=FALSE}
# Uploading csv files
q2_2019 <- read.csv("Divvy_Trips_2019_Q2.csv")
q3_2019 <- read.csv("Divvy_Trips_2019_Q3.csv")
q4_2019 <- read.csv("Divvy_Trips_2019_Q4.csv")
q1_2020 <- read.csv("Divvy_Trips_2020_Q1.csv")

# Viewing column names and structure

str(q2_2019)
colnames(q2_2019)

str(q3_2019)
colnames(q3_2019)

str(q4_2019)
colnames(q4_2019)

str(q1_2020)
colnames(q1_2020)

```

## Data Cleaning and Wrangling
This section involves renaming columns to ensure uniformity, ensuring data types are consistent, merging data sets and removing irrelevant columns and bad data. Additionally, for data aggregation, new columns would be created.

### Renaming Columns

```{r echo=TRUE, message=FALSE, warning=FALSE}
#Columns in q2_2019, q3_2019 & q4_2019 to be renamed to make them consistent with q1_2020
q2_2019 <- rename(q2_2019,
                 ride_id = "X01...Rental.Details.Rental.ID",
                 rideable_type = "X01...Rental.Details.Bike.ID",
                 started_at = "X01...Rental.Details.Local.Start.Time",
                 ended_at = "X01...Rental.Details.Local.End.Time",
                 start_station_name = "X03...Rental.Start.Station.Name",
                 start_station_id = "X03...Rental.Start.Station.ID",
                 end_station_name = "X02...Rental.End.Station.Name",
                 end_station_id = "X02...Rental.End.Station.ID",
                 member_casual = "User.Type")

q3_2019 <- rename(q3_2019,
                  ride_id = "trip_id",
                  rideable_type = "bikeid",
                  started_at = "start_time",
                  ended_at = "end_time",
                  start_station_name = "from_station_name",
                  start_station_id = "from_station_id",
                  end_station_name = "to_station_name",
                  end_station_id = "to_station_id",
                  member_casual = "usertype")

q4_2019 <- rename(q4_2019,
                  ride_id = "trip_id",
                  rideable_type = "bikeid",
                  started_at = "start_time",
                  ended_at = "end_time",
                  start_station_name = "from_station_name",
                  start_station_id = "from_station_id",
                  end_station_name = "to_station_name",
                  end_station_id = "to_station_id",
                  member_casual = "usertype")

```

### Converting Data Types

```{r echo=TRUE, message=FALSE, warning=FALSE}
##Converting data types in q2_2019, q3_2019 & q4_2019 to match data types in q1_2020
q2_2019 <- mutate(q2_2019, ride_id = as.character(ride_id),
                  rideable_type = as.character(rideable_type))
q3_2019 <- mutate(q3_2019, ride_id = as.character(ride_id),
                  rideable_type = as.character(rideable_type))
q4_2019 <- mutate(q4_2019, ride_id = as.character(ride_id),
                  rideable_type = as.character(rideable_type))

```

### Combining Respective Data Sets

```{r echo=TRUE, message=FALSE, warning=FALSE}
bike_trips <- bind_rows(q1_2020, q2_2019, q3_2019, q4_2019)

```

### Removing Irrelevant Columns

```{r echo=TRUE, message=FALSE, warning=FALSE}
#birthyear, gender, start_lat, start_lng, end_lat, end_lng, member.gender, X05...member.details.member.birthday.year, tripduration, X01...Rental.details.duration.in.seconds.uncapped 
bike_trips <- bike_trips %>% 
  select(-c(birthyear, gender, start_lat, start_lng, end_lat, end_lng, Member.Gender, 
            "X05...Member.Details.Member.Birthday.Year", "tripduration", "X01...Rental.Details.Duration.In.Seconds.Uncapped"))

```

### Inspecting Combined Data Set

```{r echo=TRUE, message=FALSE, warning=FALSE}
summary(bike_trips)
str(bike_trips)
head(bike_trips)

```

### Ensuring consistency in member_casual column

```{r echo=TRUE, message=FALSE, warning=FALSE}
bike_trips <- bike_trips %>% 
  mutate(member_casual = recode(member_casual,
         "Subscriber" = "member",
         "Customer" = "casual"))

```

### Creating New Columns
New columns **(Day, Month & Year)** were created to enable data aggregation

```{r echo=TRUE, message=FALSE, warning=FALSE}
bike_trips$date <- as.Date(bike_trips$started_at)
bike_trips$month <- format(as.Date(bike_trips$date),"%m")
bike_trips$day <- format(as.Date(bike_trips$date),"%d")
bike_trips$year <- format(as.Date(bike_trips$date),"%Y")
bike_trips$day_of_week <- format(as.Date(bike_trips$date),"%A")

#Creating new column to calculate duration of each ride 
bike_trips$ride_length <- difftime(bike_trips$ended_at,bike_trips$started_at)

is.factor(bike_trips$ride_length)
bike_trips$ride_length <- as.numeric(as.character(bike_trips$ride_length))
is.numeric(bike_trips$ride_length)

```

### Removing Bad Data

```{r echo=TRUE, message=FALSE, warning=FALSE}
bike_trips_v2 <- bike_trips[!(bike_trips$start_station_name == "HQ QR" | bike_trips$ride_length<0),]

```

### Arranging Days of the Week in Order

```{r echo=TRUE, message=FALSE, warning=FALSE}
bike_trips_v2$day_of_week <- ordered(bike_trips_v2$day_of_week, levels=c("Sunday", "Monday", "Tuesday", "Wednesday", "Thursday", "Friday", "Saturday"))
```


# Descriptive Analysis

```{r echo=TRUE, message=FALSE, warning=FALSE}
mean(bike_trips_v2$ride_length)
median(bike_trips_v2$ride_length)
max(bike_trips_v2$ride_length)
min(bike_trips_v2$ride_length)
summary(bike_trips_v2$ride_length)

```

### Comparing Members vs Casual Riders

```{r message=FALSE, warning=FALSE}
aggregate(bike_trips_v2$ride_length ~ bike_trips_v2$member_casual, FUN = mean)
aggregate(bike_trips_v2$ride_length ~ bike_trips_v2$member_casual, FUN = median)
aggregate(bike_trips_v2$ride_length ~ bike_trips_v2$member_casual, FUN = max)
aggregate(bike_trips_v2$ride_length ~ bike_trips_v2$member_casual, FUN = min)

```

### Comparing Average Ride Times by Day for Members vs Casual Rider

```{r echo=TRUE, message=FALSE, warning=FALSE}
aggregate(bike_trips_v2$ride_length ~ bike_trips_v2$member_casual + bike_trips_v2$day_of_week, FUN = mean)

```

### Analyzing Ridership Data by Type & Weekday

```{r echo=TRUE, message=FALSE, warning=FALSE}
bike_trips_v2 %>% 
  mutate(weekday = wday(started_at, label = TRUE)) %>% 
  group_by(member_casual, weekday) %>% 
  summarise(number_of_rides = n(), average_duration = mean(ride_length)) %>% 
  arrange(member_casual, weekday)

```

# Visualizing Findings

```{r echo=TRUE, message=FALSE, warning=FALSE}
bike_trips_v2 %>%
  mutate(weekday = wday(started_at, label = TRUE)) %>%
  group_by(member_casual, weekday) %>% 
  summarise(number_of_rides = n(),
            average_duration = mean(ride_length)) %>% 
  arrange(member_casual, weekday) %>% 
  ggplot(aes(x = weekday, y = average_duration, fill = member_casual)) +
  geom_col(position = "dodge")

```

### Visualizing by Rider Type

```{r echo=TRUE, message=FALSE, warning=FALSE}
 bike_trips_v2 %>% 
    mutate(weekday = wday(started_at, label = TRUE)) %>% 
    group_by(member_casual, weekday) %>% 
    summarise(number_of_rides = n(),
              average_duration = mean(ride_length)) %>% 
    ggplot(aes(x = weekday, y = number_of_rides, fill = member_casual)) + 
    geom_col(position = "dodge")

```