CYCLISTIC BIKE SHARE - CASE STUDY WITH R

knitr::include_graphics(here::here("Bike_station.jpg"))

Bike Station

INTRODUCTION

This capstone project is the final project in my Google Data Analytics Professional Certificate Course. In this case study, I will be analyzing a public dataset for a fictional company called Cyclistic, provided by the course. Here, I will be using R programming language for this analysis because of its potential benefits to reproducibility, transparency, easy statistical analysis tools and data visualizations.

The following sets of data analysis process will be followed:

• Ask,

• Prepare,

• Process,

• Analyze,

• Share,

• Act.

The case study road map as listed below will be followed on each step

• Codes, when needed.

• Key tasks.

• Deliverables.

Scenerio

You are a junior data analyst working in the marketing analyst team at Cyclistic, a bike-share company in Chicago. The director of marketing believes the company’s future success depends on maximizing the number of annual memberships. Therefore, your team wants to understand how casual riders and annual members use Cyclistic bikes differently. From these insights, your team will design a new marketing strategy to convert casual riders into annual members.But first, Cyclistic executives must approve your recommendations, so they must be backed up with compelling data insights and professional data.

ASK

• Three questions will guide the future marketing program:

1. How do annual members and casual riders use Cyclistic bikes differently?

2. Why would casual riders buy Cyclistic annual memberships?

3. How can Cyclistic use digital media to influence casual riders to become members?

Lily Moreno (director of marketing and my manager) has assigned me the first question to answer: How do annual members and casual riders use Cyclistic bikes differently?

Key tasks

1. Identify the business task

• The main business obejective is to design marketing strategies aimed at converting casual riders into annual members by understanding how they differ.

2. Consider key stakeholders

• The key stakeholders are the Director of Marketing (Lily Moreno), Marketing Analytics team, and Executive team.

Deliverable

1. A clear statement of the business task

• To find the differences between the casual riders and annual members.

PREPARE

• I will use Cyclistic’s historical trip data to analyze and identify trends.The data has been made available by Motivate International Inc. under this license. Datasets are available here link.

Key tasks

1. Download data and store it appropriately.

• Data has been downloaded and copies have been stored securely on my computer.

2. Identify how it’s organized.

• The data is in CSV (comma-separated values) format, and there are a total of 13 columns.

3. Sort and filter the data.

• For this analysis, I will be using data for the year 2019 and 2020.

4. Determine the credibility of the data.

• For the purposes of this case study, the datasets are appropriate and will enable me to answer the business questions. The data has been made available by Motivate International Inc. This is public data that I can use to explore how different customer types are using Cyclistic bikes. But data-privacy issues will prohibit me from using rider’s personally identifiable information and this will prevent me from determining if riders have purchased multiple single passes. All ride ids are unique.

Deliverable

1. A description of all data sources used

• The main source of all the data used was provided by the Cyclistic Company.

Install and load required packages

install.packages("tidyverse")
library(tidyverse)

install.packages("lubridate")
library(lubridate)

install.packages("ggplot2")
library(ggplot2)

Import data to R Studio

• Importing data for 2019

read_csv("Divvy_Trips_2019_Q1.csv")
q1_2019 <- read_csv("Divvy_Trips_2019_Q1.csv")
read_csv("Divvy_Trips_2019_Q2.csv")
q2_2019 <- read_csv("Divvy_Trips_2019_Q2.csv")
read_csv("Divvy_Trips_2019_Q3.csv")
q3_2019 <- read_csv("Divvy_Trips_2019_Q3.csv")
read_csv("Divvy_Trips_2019_Q4.csv")
q4_2019 <- read_csv("Divvy_Trips_2019_Q4.csv")

• Importing data for 2020

read_csv("Divvy_Trips_2020_Q1.csv")
q1_2020 <- read_csv("Divvy_Trips_2020_Q1.csv")

read_csv("202004-divvy-tripdata.csv")
q2_04 <- read_csv("202004-divvy-tripdata.csv")
read_csv("202005-divvy-tripdata.csv")
q2_05 <- read_csv("202005-divvy-tripdata.csv")
read_csv("202006-divvy-tripdata.csv")
q2_06 <- read_csv("202006-divvy-tripdata.csv")

bind_rows(q2_04, q2_05, q2_06)
q2_2020 <- bind_rows(q2_04, q2_05, q2_06)

read_csv("202007-divvy-tripdata.csv")
q3_07 <- read_csv("202004-divvy-tripdata.csv")
read_csv("202008-divvy-tripdata.csv")
q3_08 <- read_csv("202008-divvy-tripdata.csv")
read_csv("202009-divvy-tripdata.csv")
q3_09 <- read_csv("202009-divvy-tripdata.csv")
bind_rows(q3_07, q3_08, q3_09)
q3_2020 <- bind_rows(q3_07, q3_08, q3_09)

read_csv("202010-divvy-tripdata.csv")
q4_10 <- read_csv("202010-divvy-tripdata.csv")
read_csv("202011-divvy-tripdata.csv")
q4_11 <- read_csv("202011-divvy-tripdata.csv")
read_csv("202012-divvy-tripdata.csv")
q4_12 <- read_csv("202012-divvy-tripdata.csv")
rbind(q4_10, q4_11, q4_12)
q4_2020 <- rbind(q4_10, q4_11, q4_12)

Wrangle and merge all data into a single file

• I made sure all the data sets have the same number of columns and also the name of each columns are the same before merging it all together

bind_rows(q1_2019, q2_2019, q3_2019, q4_2019
          , q1_2020, q2_2020, q3_2020, q4_2020)

all_trips <- bind_rows(q1_2019, q2_2019, q3_2019, q4_2019
                       , q1_2020, q2_2020, q3_2020, q4_2020)

PROCESS

Cleaning up data and adding data to prepare for analysis

Key tasks

1. Check the data for errors.

2. Choose your tools.

3. Transform the data so you can work with it effectively.

4. Document the cleaning process.

Deliverables

1. Documentation of any cleaning or manipulation of data.

• I inspected the new table that has been created using the following code chunks

colnames(all_trips)  #List of column names
nrow(all_trips)  #How many rows are in data frame
dim(all_trips)  #Dimensions of the data frame
head(all_trips)  #See the first 6 rows of data frame
tail(all_trips)  #see the last 6 rows of data frame
str(all_trips)  #See list of columns and data types (numeric, character, etc)
summary(all_trips)  #Statistical summary of data. Mainly for numeric

Adding columns that list the date, month, day, and year of each ride.

• This will allow us to aggregate ride data for each month, day, or year. Therefore, the code chunks used are as follows

all_trips$date <- as.Date(all_trips$started_at) #The default format is yyyy-mm-dd
all_trips$month <- format(as.Date(all_trips$date), "%m")
all_trips$day <- format(as.Date(all_trips$date), "%d")
all_trips$year <- format(as.Date(all_trips$date), "%Y")
all_trips$day_of_week <- format(as.Date(all_trips$date), "%A")

Adding a “ride_length” calculation to all_trips (in seconds)

all_trips$ride_length <- difftime(all_trips$ended_at,all_trips$started_at)

str(all_trips)  #to inspect the structure of the columns

Convert “ride_length” from factor to numeric so we can run calculations on the data

all_trips$ride_length <- as.numeric(as.character(all_trips$ride_length))

is.numeric(all_trips$ride_length)

Remove “bad” data

• The dataframe includes a few hundred entries when bikes were taken out of docks and checked for quality by Divvy or ride_length was negative or Zero, so I created a new version of the dataframe (v2) since data is being removed

all_trips[!(all_trips$start_station_name == "HQ QR" | all_trips$ride_length<0),]

all_trips_v2 <- all_trips[!(all_trips$start_station_name == "HQ QR" | all_trips$ride_length<0),]

ANALYZE

• All the data have been strored appropriately and has been prepared for analysis, so they are ready for exploration.

Key tasks

1. Aggregate your data so it’s useful and accessible.
2. Organize and format your data.
3. Perform calculations.
4. Identify trends and relationships.

Deliverables

1. A summary of your analysis

Conduct descriptive analysis - Descriptive analysis on ride_length (all figures in seconds)

mean(all_trips_v2$ride_length, na.rm = TRUE)#straight average (total ride length / rides)

median(all_trips_v2$ride_length, na.rm = TRUE)#midpoint number in the ascending array of ride lengths

max(all_trips_v2$ride_length, na.rm = TRUE)#longest ride

min(all_trips_v2$ride_length, na.rm = TRUE)#shortest_ride

knitr::include_graphics(here::here("mean median max min of all_trips.png"))

mean median max min of all_trips

Let’s visualize members and casuals by the total ride taken (ride count)

all_trips_v2 %>% 
  group_by(member_casual) %>% 
  summarise(ride_count = length(ride_id)) %>%
  ggplot(aes(x = member_casual, y = ride_count, fill = member_casual)) +
  geom_col(position = "dodge") +
  labs(title ="Total rides taken (ride_count) of Members and Casual riders") +
  geom_col(width=0.5, position = position_dodge(width=0.5)) +
  scale_y_continuous(labels = function(x) format(x, scientific = FALSE))

knitr::include_graphics(here::here("member_casual ride_count.png"))

member_casual ride_count

knitr::include_graphics(here::here("Total rides taken (ride_count) of Members and Casual riders.png"))

• From the above graph, we can observe that there are more member riders compared to casual rides based on the ride count.

Let’s see the average time ride by each day for members vs casual users

aggregate(all_trips_v2$ride_length ~ all_trips_v2$member_casual 
          + all_trips_v2$day_of_week
          , FUN = mean)

# notice that the above code result didn't give us a well ordered days of the week
# now let's put it in a well arranged order

all_trips_v2$day_of_week <- ordered(all_trips_v2$day_of_week, levels=c("Sunday"
                                                                       , "Monday"
                                                                       , "Tuesday"
                                                                       , "Wednesday"
                                                                       , "Thursday"
                                                                       , "Friday"
                                                                       , "Saturday"))

#now we'd rerun the average time ride by each day for members vs casual users
# so as to see if the ordered code we entered would work
aggregate(all_trips_v2$ride_length ~ all_trips_v2$member_casual 
          + all_trips_v2$day_of_week
          , FUN = mean)

knitr::include_graphics(here::here("average ride time by each day for member vs casual users.png"))

average ride time by each day for member vs casual users

all_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)   

knitr::include_graphics(here::here("total rides and average ride time by each day for members vs casual riders.png"))

total rides and average ride time(duration) by each day for members vs casual riders

Let’s visualize the above table by days of the week and number of rides taken by member and casual riders.

knitr::include_graphics(here::here("Total rides of Members and Casual riders Vs. Day of the week.png"))

• From the above graph, it is observed that the members are quite consistent with higher number of rides throughout the week compared to the casual riders, and also, the differences in number of rides between the members and casual riders during the weekends(Saturdays and Sundays) is not as much as the differences during the other days of the week.

Let’s visualize the average duration of Members and Casual riders Vs. Day of the week

all_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") +
  labs(title ="Average duration of Members and Casual riders Vs. Day of the week") +
  geom_col(width=0.5, position = position_dodge(width=0.5)) +
  scale_y_continuous(labels = function(x) format(x, scientific = FALSE))

knitr::include_graphics(here::here("Average duration of Members and Casual riders Vs. Day of the week.png"))

• From the graph above, it is observed that the casual riders ride for a longer period of time throughout the week while the members ride at a consistent pace during the week with the highest rides on the weekend.

Let’s create a visualization for Total rides by members and casual riders by month

all_trips_v2 %>%  
  group_by(member_casual, month) %>% 
  summarise(number_of_rides = n(),.groups="drop") %>% 
  arrange(member_casual, month)  %>% 
  ggplot(aes(x = month, y = number_of_rides, fill = member_casual)) +
  labs(title ="Total rides by Members and Casual riders by Month") +
  theme(axis.text.x = element_text(angle = 45)) +
  geom_col(width=0.5, position = position_dodge(width=0.5)) +
  scale_y_continuous(labels = function(x) format(x, scientific = FALSE))

knitr::include_graphics(here::here("Total rides by Members and Casual riders by Month.png"))

• From the above graph, it is observed that members have the highest number of rides throughout the year with August being the month with the highest number of rides in general for both members and casual riders.

Let’s compare Members and Casual riders depending on ride distance.

all_trips_v2 %>% 
  group_by(member_casual) %>% drop_na() %>%
  summarise(average_ride_length = mean(ride_length)) %>%
  ggplot() + 
  geom_col(mapping= aes(x= member_casual,y= average_ride_length,fill=member_casual), show.legend = FALSE)+
  labs(title = "Mean distance traveled by Members and Casual riders")

knitr::include_graphics(here::here("Mean distance traveled by Members and Casual riders.png"))

• From the graph above, we can observe that the distance traveled by the casual riders is far more than the distance traveled by the member with a very large difference in kilometers.

SHARE

This phase involves using visualization to share my findings and can be done by presentation.

Key tasks

1. Determine the best way to share your findings.

2. Create effective data visualizations.

3. Present your findings.

4. Ensure your work is accessible.

Deliverables

1. Supporting visualizations and key findings

ACT

This phase will be carried out by the executive team, Director of Marketing (Lily Moreno) and the Marketing Analytics team based on my analysis.

Conclusion

1. Members have more bikes compared to casual riders.

2. We have more members riding in all months compared to casual riders.

3. Casual riders travel for a longer time period.

4. Members ride more throughout the entire weekday while the casual riders also have a high ride record during the weekends(Saturday and Sunday) compared to the other days of the week.

5. Casual riders go farther in terms of distance.

Deliverable

• Your top 3 recommendations based on your analysis

1. Have a slash sale or promo for casual riders so they can acquire more bikes and indulge them in the benefits of being a member.

2. Host fun biking competitions with prizes at intervals for casual riders on the weekends. Since there are lot of members on weekends,this will also attract them to get a membership.

3. Encourage casual riders to ride more in the entire year through advertisement, hand flyers, by giving them various coupons so as to convince them into being a member.

THANK YOU FOR READING, PLEASE PROVIDE YOUR VALUABLE FEEDBACK.