Google-Data-Analytics-Capstone-Bike-Share

Cyclistic Bike Sharing Analysis

Loading the libraaries required for installing, Cleansing and Visualization

library(tidyverse) # metapackage of all tidyverse packages library(dplyr) library(geosphere) library(lubridate) library(ggplot2)

Uploading and Saving Dataset

aug <- read_csv("../input/august/202208-divvy-tripdata.csv") jul <- read_csv("../input/july22/202207-divvy-tripdata.csv") jun <- read_csv("../input/june22/202206-divvy-tripdata.csv") may <- read_csv("../input/may2022/202205-divvy-tripdata.csv") mar <- read_csv("../input/march22/202203-divvy-tripdata.csv") feb <- read_csv("../input/feb2022/202202-divvy-tripdata.csv") jan <- read_csv("../input/jan2022/202201-divvy-tripdata.csv") dec <- read_csv("../input/dec2021/202112-divvy-tripdata.csv") nov <- read_csv("../input/nov2021/202111-divvy-tripdata.csv") oct <- read_csv("../input/oct2022/202110-divvy-tripdata.csv") sep <- read_csv("../input/sep2022/202109-divvy-tripdata.csv")

Converting and unifying datatype

aug <- mutate(aug, start_station_id = as.character(start_station_id) ,end_station_id = as.character(end_station_id) ,started_at= as.POSIXct(started_at, format= "%m/%d/%Y %H:%M") ,ended_at= as.POSIXct(ended_at, format= "%m/%d/%Y %H:%M"))

jul <- mutate(jul, start_station_id = as.character(start_station_id) ,end_station_id = as.character(end_station_id) ,started_at= as.POSIXct(started_at, format= "%m/%d/%Y %H:%M") ,ended_at= as.POSIXct(ended_at, format= "%m/%d/%Y %H:%M"))

jun <- mutate(jun, start_station_id = as.character(start_station_id) ,end_station_id = as.character(end_station_id) ,started_at= as.POSIXct(started_at, format= "%m/%d/%Y %H:%M") ,ended_at= as.POSIXct(ended_at, format= "%m/%d/%Y %H:%M"))

may <- mutate(may, start_station_id = as.character(start_station_id) ,end_station_id = as.character(end_station_id) ,started_at= as.POSIXct(started_at, format= "%m/%d/%Y %H:%M") ,ended_at= as.POSIXct(ended_at, format= "%m/%d/%Y %H:%M"))

mar <- mutate(mar, start_station_id = as.character(start_station_id) ,end_station_id = as.character(end_station_id) ,started_at= as.POSIXct(started_at, format= "%m/%d/%Y %H:%M") ,ended_at= as.POSIXct(ended_at, format= "%m/%d/%Y %H:%M"))

feb <- mutate(feb, start_station_id = as.character(start_station_id) ,end_station_id = as.character(end_station_id) ,started_at= as.POSIXct(started_at, format= "%m/%d/%Y %H:%M") ,ended_at= as.POSIXct(ended_at, format= "%m/%d/%Y %H:%M"))

jan <- mutate(jan, start_station_id = as.character(start_station_id) ,end_station_id = as.character(end_station_id) ,started_at= as.POSIXct(started_at, format= "%m/%d/%Y %H:%M") ,ended_at= as.POSIXct(ended_at, format= "%m/%d/%Y %H:%M"))

dec <- mutate(dec, start_station_id = as.character(start_station_id) ,end_station_id = as.character(end_station_id) ,started_at= as.POSIXct(started_at, format= "%m/%d/%Y %H:%M") ,ended_at= as.POSIXct(ended_at, format= "%m/%d/%Y %H:%M"))

nov <- mutate(nov, start_station_id = as.character(start_station_id) ,end_station_id = as.character(end_station_id) ,started_at= as.POSIXct(started_at, format= "%m/%d/%Y %H:%M") ,ended_at= as.POSIXct(ended_at, format= "%m/%d/%Y %H:%M"))

oct <- mutate(oct, start_station_id = as.character(start_station_id) ,end_station_id = as.character(end_station_id) ,started_at= as.POSIXct(started_at, format= "%m/%d/%Y %H:%M") ,ended_at= as.POSIXct(ended_at, format= "%m/%d/%Y %H:%M"))

sep <- mutate(sep, start_station_id = as.character(start_station_id) ,end_station_id = as.character(end_station_id) ,started_at= as.POSIXct(started_at, format= "%m/%d/%Y %H:%M") ,ended_at= as.POSIXct(ended_at, format= "%m/%d/%Y %H:%M"))

apr <- mutate(apr, start_station_id = as.character(start_station_id) ,end_station_id = as.character(end_station_id) ,started_at= as.POSIXct(started_at, format= "%m/%d/%Y %H:%M") ,ended_at= as.POSIXct(ended_at, format= "%m/%d/%Y %H:%M"))

Adding all the files to merge

bikedata <- bind_rows(aug,jul,jun,may,mar,feb,jan,dec,nov)

Clean dataset by removing unnessesary colons

ridedata <- ridedata %>% select(-c(start_station_id, end_station_id, start_lat, start_lng, end_lat, end_lng))

inspecting the new file

colnames(bikedata) str(bikedata) head(ridedata)

Adding all the files to merge

bikedata1 <- bind_rows(aug,jul,jun,may,apr,mar,feb,jan,dec,nov)

Clean dataset by removing unnessesary colons

bikedata2 <- bikedata1 %>% select(-c(start_station_id, end_station_id, start_lat, start_lng, end_lat, end_lng))

inspecting the new file

colnames(bikedata2) str(bikedata2) head(bikedata2)

Add date, month, day and trip length columns all at the same time

bikedata2$date <- as.Date(bikedata2$started_at) bikedata2$month <- format(as.Date(bikedata2$date), "%B") bikedata2$day <- format(as.Date(bikedata2$date), "%A") bikedata2$trip_length <- difftime(bikedata2$ended_at, bikedata2$started_at)

remove empty rows, dataset trimmed down from3489748 rows to 3294691

colSums(is.na(bikedata2)) #returns a summary of empty colons on each variable cleanbikedata2 <- bikedata2 [complete.cases(bikedata2), ]

Remove all negative and zero trip length as well as NA/null values

cleanbikedata2 <- subset(cleanbikedata2, trip_length > 0)

Convert ride length to integer

cleanbikedata2$trip_length <- as.integer(cleanbikedata2$trip_length)

DESCRIPTIVE ANALYSIS

Aggregating Data to determine average trip length of riders for month of the year

aggregate(trip_length ~ member_casual + month, cleanbikedata2, mean)

Aggregate data to determine average trip length of riders on each day of the week

aggregate(trip_length ~ day + member_casual, cleanbikedata2, mean)

Aggregate data to determine how riders use the different bike types available

aggregate(trip_length ~ rideable_type + member_casual, cleanbikedata2, mean)

DATA VISUALIZATION

1.Group and visualize data by rider type and monthly trip length

mc_month <- cleanbikedata2 %>% group_by(month, member_casual) %>% count() ggplot(data=mc_month)+ geom_col(mapping=aes(x=month, y=n, fill=member_casual))+ facet_wrap(~member_casual)+ labs(y="Trip length", x="Rider type", fill="Member/Casual", title="Ride length by month")+ theme(axis.text.x=element_text(angle=45))

2.Group and visualize rider type by day of the week

mc_week <- cleanbikedata2 %>% group_by(day, member_casual) %>% count() ggplot(data=mc_week)+ geom_col(mapping=aes(x=day, y=n, fill=member_casual))+ facet_grid(~member_casual)+ labs(y="Trip length", x="Rider type", fill="Member/Casual", title="Ride length by day of the week")+ theme(axis.text.x=element_text(angle=45))

3.Group, summarize and visualize data to determine ride count by rider type

mcdata <- cleanbikedata2 %>% group_by(member_casual) %>% count() ggplot(data=mcdata)+ geom_col(mapping=aes(x=member_casual, y=n, fill=member_casual))+ labs(y="Trip length", x="Rider type", fill="Member/Casual", title="Number of trips by different rider types")

4.Summarize and visualize data by rider type and average trip length to ascertain lenght of ride taken by different rider types

trip_len <- cleanbikedata2 %>% group_by(member_casual) %>% summarise(mean_trip_length = mean(trip_length, na.rm = TRUE)) ggplot(data=trip_len)+ geom_col(mapping=aes(x=member_casual, y=mean_trip_length, fill=member_casual))+ labs(y="Trip length", x="Rider type", fill="Member/Casual", title="Lenght of ride taken by different rider types")

5.Group and visualize data to determine prefered type of bike by riders

Group and visualize data by member_casual and rideable_type to determine prefered type of bike by riders

ride_type <- cleanbikedata2 %>% group_by(member_casual, rideable_type) %>% count() ggplot(data=ride_type)+ geom_col(mapping=aes(x=rideable_type, y=n, fill=member_casual))+ facet_wrap(~member_casual)+ labs(y="Number of rides", x="Type of bike", fill="Member/Casual", title="Most prefered type of bike by riders")

TOP 10 STATIONS

Creat a new data frame for all stations

all_stations <- bind_rows(data.frame("stations" = cleanbikedata2$start_station_name, "member_casual" = cleanbikedata2$member_casual), data.frame("stations" = cleanbikedata2$end_station_name, "member_casual" = cleanbikedata2$member_casual))

Exclude entries with no station name

all_stations_v2 <- all_stations[!(all_stations$stations == "" | is.na(all_stations$stations)),]

Separate the data frame by rider type

all_stations_member <- all_stations_v2[all_stations_v2$member_casual == 'member',] all_stations_casual <- all_stations_v2[all_stations_v2$member_casual == 'casual',]

Get the top 10 popular stations all, members and casual riders

top_10_station <- all_stations_v2 %>% group_by(stations) %>% summarise(station_count = n()) %>% arrange(desc(station_count)) %>% slice(1:10)

Viz

ggplot(data=top_10_station)+ geom_col(mapping=aes(x=stations,y=station_count, fill=stations))+ labs(title="Top 10 stations", subtitle = "Most populous bike stations.")+ theme(axis.text = element_text(angle=45))

Get the top 10 popular stations for members

top_10_stations_member <- all_stations_member %>% group_by(stations) %>% summarise(station_count = n()) %>% arrange(desc(station_count)) %>% head(n=10)

Viz

ggplot(data=top_10_stations_member)+ geom_col(mapping=aes(x=stations,y=station_count, fill=stations))+ labs(title="Top 10 stations", subtitle = "Most populous bike stations amongst members.")+ theme(axis.text = element_text(angle=45))

Get the top 10 popular stations for members

top_10_stations_casual <- all_stations_casual %>% group_by(stations) %>% summarise(station_count = n()) %>% arrange(desc(station_count)) %>% head(n=10)

Viz

ggplot(data=top_10_stations_casual)+ geom_col(mapping=aes(x=stations,y=station_count, fill=stations))+ labs(title="Top 10 stations", subtitle = "Most populous bike stations amongst members.")+ theme(axis.text = element_text(angle=45))