Uber vs. Lyft In NYC

library(stringr)
library(ggplot2)
library(scales) # y axis
library(rgdal)
library(tidyverse)
library(lubridate) # convert time to 24 hours,convert days to weekdays
library(geojsonR) # read geojson file
library(geojsonio)

Analysis of Annual Financial Reports

Lyft Annual Report

Uber Annual Report

Revenue

# in million dollars
Year = c(2016:2021)
 
Uber_Revenue = c(3338,7402,10433,13000,11139,17455)
Lyft_Revenue = c(343,1060,2157,3616,2365,3208)
Revenue = data.frame(Year,Uber_Revenue,Lyft_Revenue)

Create revenue change percentage columns

Revenue$Uber_Revenue_change_perc = c(NA,100*diff(Revenue$Uber_Revenue)/Revenue$Uber_Revenue[c(1:5)])
Revenue$Lyft_Revenue_change_perc = c(NA,100*diff(Revenue$Lyft_Revenue)/Revenue$Lyft_Revenue[c(1:5)])

Revenue_gather1 = gather(Revenue,key = "app_taxi",value = "Revenue",2:3)
Revenue_gather2 = gather(Revenue, key = "percentage", value = "perc",4:5)
Revenue_gather1$order = c(1:12)
Revenue_gather2$order = c(1:12)
Revenue_gather = merge(Revenue_gather1[,c(1,4,5,6)],Revenue_gather2[,c(4:6)],by = "order")

ggplot(Revenue_gather) + 
  aes(x = Year, y = Revenue, fill = app_taxi) +
  geom_col() +
  scale_x_continuous(breaks = 2016:2021)+
  geom_text(aes(x = Year, 
                y = Revenue, 
                label = ifelse(is.na(perc), "",
                               paste(round(perc,1),"%"))),
            vjust = -0.5) +
  facet_wrap(~app_taxi) +
  labs(x = "",
       y = "Total Revenue (In Million Dollars)",
       title = "Total Revenue and Percentage Change
       Worldwide")

Active Riders

Year = c(2016:2021)
Uber_Active_Riders = c(45,68,91,111,93,118)
Lyft_Active_Riders = c(6.6,12.6,18.6,22.905,12.552,18.728)
consumers = data.frame(Year,Uber_Active_Riders,Lyft_Active_Riders)

consumers$Uber_Rider_change_perc = c(NA,100*diff(consumers$Uber_Active_Riders)/consumers$Uber_Active_Riders[c(1:5)])
consumers$Lyft_Rider_change_perc = c(NA,100*diff(consumers$Lyft_Active_Riders)/consumers$Lyft_Active_Riders[c(1:5)])

consumers_gather1 = gather(consumers,key = "app_taxi",value = "Riders",2:3)
consumers_gather2 = gather(consumers, key = "percentage", value = "perc",4:5)
consumers_gather1$order = c(1:12)
consumers_gather2$order = c(1:12)
consumers_gather = merge(consumers_gather1[,c(1,4,5,6)],consumers_gather2[,c(4:6)],by = "order")

ggplot(consumers_gather) + 
  aes(x = Year, y = Riders, fill = app_taxi) +
  geom_col() +
  scale_x_continuous(breaks = 2016:2021)+
  geom_text(aes(x = Year, 
                y = Riders, 
                label = ifelse(is.na(perc), "",paste(round(perc,1),"%"))),
            vjust = -0.5) +
  facet_wrap(~app_taxi) +
  labs(x = "",
       y = "Active Riders (In Millions)",
       title = "Uber and Lyft Active Riders and Percentage Change
       Worldwide")

Trips

Uber_trips = c(1818,3736,5220,6904,5025,6368)
Lyft_trips = c(162.6,375.5,619.4,2100,NA,NA)
Trips = data.frame(Year,Uber_trips，Lyft_trips)

Trips$Uber_Trip_change_perc = c(NA,100*diff(Trips$Uber_trips)/Trips$Uber_trips[c(1:5)])
Trips$Lyft_Trip_change_perc = c(NA, 100*diff(Trips$Lyft_trips)/Trips$Lyft_trips[c(1:4)])

Trips_gather1 = gather(Trips,key = "app_taxi",value = "Rides",2:3)
Trips_gather2 = gather(Trips, key = "percentage", value = "perc",4:5)
Trips_gather1$order = c(1:12)
Trips_gather2$order = c(1:12)
Trips_gather = merge(Trips_gather1[,c(1,4,5,6)],Trips_gather2[,c(4:6)],by = "order")

ggplot(Trips_gather) + 
  aes(x = Year, y = Rides, fill = app_taxi) +
  geom_col() +
  scale_x_continuous(breaks = 2016:2021)+
  geom_text(aes(x = Year, 
                y = Rides, 
                label = ifelse(is.na(perc), "",paste(round(perc,1),"%"))),
            vjust = -0.5) +
  facet_wrap(~app_taxi) +
  labs(x = "",
       y = "Trips (In Millions)",
       title = "Uber and Lyft Trips and Percentage Change
       Worldwide")

Analysis of 2020_High_Volume_FHV_Trip_Records dataset

2020 High Volume FHV Trip Records

FHV = read.csv("2020_High_Volume_FHV_Trip_Records.csv")

FHV_list = vector(mode = "list", length = 14)
FHV_list

#split the dataset into smaller datasets
FHV_list[[1]] = FHV[1:10000000,]
FHV_list[[2]] = FHV[10000001:20000000,]
FHV_list[[3]] = FHV[20000001:30000000,]
FHV_list[[4]] = FHV[30000001:40000000,]
FHV_list[[5]] = FHV[40000001:50000000,]
FHV_list[[6]] = FHV[50000001:60000000,]

FHV_list[[7]] = FHV[60000001:70000000,]
FHV_list[[8]] = FHV[70000001:80000000,]
FHV_list[[9]] = FHV[80000001:90000000,]
FHV_list[[10]] = FHV[90000001:100000000,]
FHV_list[[11]] = FHV[100000001:110000000,]
FHV_list[[12]] = FHV[110000001:120000000,]
FHV_list[[13]] = FHV[120000001:130000000,]
FHV_list[[14]] = FHV[130000001:nrow(FHV),]

Each subset is about one gigabyte. It takes time to perform one single step. I just change FHV_list[[1]], to FHV_list[[2]], FHV_list[[3]]…FHV_list[[14]] manually, and run all the steps, then I got 14 subsets.

FHV_list[[1]]$pickup_datetime = as.POSIXct(FHV_list[[1]]$pickup_datetime, format = "%m/%d/%Y %I:%M:%S %p",tz = Sys.timezone())

FHV_list[[1]]$dropoff_datetime = as.POSIXct(FHV_list[[1]]$dropoff_datetime, format = "%m/%d/%Y %I:%M:%S %p",tz = Sys.timezone())

# create Year, Months, Time, Trip_duration and Hours columns
FHV_list[[1]]$Year = format(FHV_list[[1]]$pickup_datetime, format = "%Y")

FHV_list[[1]]$Months = format(FHV_list[[1]]$pickup_datetime, format = "%m")

FHV_list[[1]]$Time = format(FHV_list[[1]]$pickup_datetime, format = "%H:%M:%S")

FHV_list[[1]]$Trip_duration = difftime(FHV_list[[1]]$dropoff_datetime, FHV_list[[1]]$pickup_datetime, units = "hours")

FHV_list[[1]]$Hours = format(FHV_list[[1]]$pickup_datetime, format = "%H")

FHV_list[[1]]$Weekdays = weekdays(FHV_list[[1]]$pickup_datetime)

FHV_list[[2]]$pickup_datetime = as.POSIXct(FHV_list[[2]]$pickup_datetime, format = "%m/%d/%Y %I:%M:%S %p",tz = Sys.timezone())

FHV_list[[2]]$dropoff_datetime = as.POSIXct(FHV_list[[2]]$dropoff_datetime, format = "%m/%d/%Y %I:%M:%S %p",tz = Sys.timezone())

# create Year, Months, Time, Trip_duration and Hours columns
FHV_list[[2]]$Year = format(FHV_list[[2]]$pickup_datetime, format = "%Y")

FHV_list[[2]]$Months = format(FHV_list[[2]]$pickup_datetime, format = "%m")

FHV_list[[2]]$Time = format(FHV_list[[2]]$pickup_datetime, format = "%H:%M:%S")

FHV_list[[2]]$Trip_duration = difftime(FHV_list[[2]]$dropoff_datetime, FHV_list[[2]]$pickup_datetime, units = "hours")

FHV_list[[2]]$Hours = format(FHV_list[[2]]$pickup_datetime, format = "%H")

FHV_list[[2]]$Weekdays = weekdays(FHV_list[[2]]$pickup_datetime)

FHV_list[[3]]$pickup_datetime = as.POSIXct(FHV_list[[3]]$pickup_datetime, format = "%m/%d/%Y %I:%M:%S %p",tz = Sys.timezone())

FHV_list[[3]]$dropoff_datetime = as.POSIXct(FHV_list[[3]]$dropoff_datetime, format = "%m/%d/%Y %I:%M:%S %p",tz = Sys.timezone())

# create Year, Months, Time, Trip_duration and Hours columns
FHV_list[[3]]$Year = format(FHV_list[[3]]$pickup_datetime, format = "%Y")

FHV_list[[3]]$Months = format(FHV_list[[3]]$pickup_datetime, format = "%m")

FHV_list[[3]]$Time = format(FHV_list[[3]]$pickup_datetime, format = "%H:%M:%S")

FHV_list[[3]]$Trip_duration = difftime(FHV_list[[3]]$dropoff_datetime, FHV_list[[3]]$pickup_datetime, units = "hours")

FHV_list[[3]]$Hours = format(FHV_list[[3]]$pickup_datetime, format = "%H")

FHV_list[[3]]$Weekdays = weekdays(FHV_list[[3]]$pickup_datetime)

FHV_list[[4]]$pickup_datetime = as.POSIXct(FHV_list[[4]]$pickup_datetime, format = "%m/%d/%Y %I:%M:%S %p",tz = Sys.timezone())

FHV_list[[4]]$dropoff_datetime = as.POSIXct(FHV_list[[4]]$dropoff_datetime, format = "%m/%d/%Y %I:%M:%S %p",tz = Sys.timezone())

# create Year, Months, Time, Trip_duration and Hours columns
FHV_list[[4]]$Year = format(FHV_list[[4]]$pickup_datetime, format = "%Y")

FHV_list[[4]]$Months = format(FHV_list[[4]]$pickup_datetime, format = "%m")

FHV_list[[4]]$Time = format(FHV_list[[4]]$pickup_datetime, format = "%H:%M:%S")

FHV_list[[4]]$Trip_duration = difftime(FHV_list[[4]]$dropoff_datetime, FHV_list[[4]]$pickup_datetime, units = "hours")

FHV_list[[4]]$Hours = format(FHV_list[[4]]$pickup_datetime, format = "%H")

FHV_list[[4]]$Weekdays = weekdays(FHV_list[[4]]$pickup_datetime)

FHV_list[[5]]$pickup_datetime = as.POSIXct(FHV_list[[5]]$pickup_datetime, format = "%m/%d/%Y %I:%M:%S %p",tz = Sys.timezone())

FHV_list[[5]]$dropoff_datetime = as.POSIXct(FHV_list[[5]]$dropoff_datetime, format = "%m/%d/%Y %I:%M:%S %p",tz = Sys.timezone())

# create Year, Months, Time, Trip_duration and Hours columns
FHV_list[[5]]$Year = format(FHV_list[[5]]$pickup_datetime, format = "%Y")

FHV_list[[5]]$Months = format(FHV_list[[5]]$pickup_datetime, format = "%m")

FHV_list[[5]]$Time = format(FHV_list[[5]]$pickup_datetime, format = "%H:%M:%S")

FHV_list[[5]]$Trip_duration = difftime(FHV_list[[5]]$dropoff_datetime, FHV_list[[5]]$pickup_datetime, units = "hours")

FHV_list[[5]]$Hours = format(FHV_list[[5]]$pickup_datetime, format = "%H")

FHV_list[[5]]$Weekdays = weekdays(FHV_list[[5]]$pickup_datetime)

FHV_list[[6]]$pickup_datetime = as.POSIXct(FHV_list[[6]]$pickup_datetime, format = "%m/%d/%Y %I:%M:%S %p",tz = Sys.timezone())

FHV_list[[6]]$dropoff_datetime = as.POSIXct(FHV_list[[6]]$dropoff_datetime, format = "%m/%d/%Y %I:%M:%S %p",tz = Sys.timezone())

# create Year, Months, Time, Trip_duration and Hours columns
FHV_list[[6]]$Year = format(FHV_list[[6]]$pickup_datetime, format = "%Y")

FHV_list[[6]]$Months = format(FHV_list[[6]]$pickup_datetime, format = "%m")

FHV_list[[6]]$Time = format(FHV_list[[6]]$pickup_datetime, format = "%H:%M:%S")

FHV_list[[6]]$Trip_duration = difftime(FHV_list[[6]]$dropoff_datetime, FHV_list[[6]]$pickup_datetime, units = "hours")

FHV_list[[6]]$Hours = format(FHV_list[[6]]$pickup_datetime, format = "%H")

FHV_list[[6]]$Weekdays = weekdays(FHV_list[[6]]$pickup_datetime)

FHV_list[[7]]$pickup_datetime = as.POSIXct(FHV_list[[7]]$pickup_datetime, format = "%m/%d/%Y %I:%M:%S %p",tz = Sys.timezone())

FHV_list[[7]]$dropoff_datetime = as.POSIXct(FHV_list[[7]]$dropoff_datetime, format = "%m/%d/%Y %I:%M:%S %p",tz = Sys.timezone())

# create Year, Months, Time, Trip_duration and Hours columns
FHV_list[[7]]$Year = format(FHV_list[[7]]$pickup_datetime, format = "%Y")

FHV_list[[7]]$Months = format(FHV_list[[7]]$pickup_datetime, format = "%m")

FHV_list[[7]]$Time = format(FHV_list[[7]]$pickup_datetime, format = "%H:%M:%S")

FHV_list[[7]]$Trip_duration = difftime(FHV_list[[7]]$dropoff_datetime, FHV_list[[7]]$pickup_datetime, units = "hours")

FHV_list[[7]]$Hours = format(FHV_list[[7]]$pickup_datetime, format = "%H")

FHV_list[[7]]$Weekdays = weekdays(FHV_list[[7]]$pickup_datetime)

FHV_list[[8]]$pickup_datetime = as.POSIXct(FHV_list[[8]]$pickup_datetime, format = "%m/%d/%Y %I:%M:%S %p",tz = Sys.timezone())

FHV_list[[8]]$dropoff_datetime = as.POSIXct(FHV_list[[8]]$dropoff_datetime, format = "%m/%d/%Y %I:%M:%S %p",tz = Sys.timezone())

# create Year, Months, Time, Trip_duration and Hours columns
FHV_list[[8]]$Year = format(FHV_list[[8]]$pickup_datetime, format = "%Y")

FHV_list[[8]]$Months = format(FHV_list[[8]]$pickup_datetime, format = "%m")

FHV_list[[8]]$Time = format(FHV_list[[8]]$pickup_datetime, format = "%H:%M:%S")

FHV_list[[8]]$Trip_duration = difftime(FHV_list[[8]]$dropoff_datetime, FHV_list[[8]]$pickup_datetime, units = "hours")

FHV_list[[8]]$Hours = format(FHV_list[[8]]$pickup_datetime, format = "%H")

FHV_list[[8]]$Weekdays = weekdays(FHV_list[[8]]$pickup_datetime)

FHV_list[[9]]$pickup_datetime = as.POSIXct(FHV_list[[9]]$pickup_datetime, format = "%m/%d/%Y %I:%M:%S %p",tz = Sys.timezone())

FHV_list[[9]]$dropoff_datetime = as.POSIXct(FHV_list[[9]]$dropoff_datetime, format = "%m/%d/%Y %I:%M:%S %p",tz = Sys.timezone())

# create Year, Months, Time, Trip_duration and Hours columns
FHV_list[[9]]$Year = format(FHV_list[[9]]$pickup_datetime, format = "%Y")

FHV_list[[9]]$Months = format(FHV_list[[9]]$pickup_datetime, format = "%m")

FHV_list[[9]]$Time = format(FHV_list[[9]]$pickup_datetime, format = "%H:%M:%S")

FHV_list[[9]]$Trip_duration = difftime(FHV_list[[9]]$dropoff_datetime, FHV_list[[9]]$pickup_datetime, units = "hours")

FHV_list[[9]]$Hours = format(FHV_list[[9]]$pickup_datetime, format = "%H")

FHV_list[[9]]$Weekdays = weekdays(FHV_list[[9]]$pickup_datetime)

FHV_list[[10]]$pickup_datetime = as.POSIXct(FHV_list[[10]]$pickup_datetime, format = "%m/%d/%Y %I:%M:%S %p",tz = Sys.timezone())

FHV_list[[10]]$dropoff_datetime = as.POSIXct(FHV_list[[10]]$dropoff_datetime, format = "%m/%d/%Y %I:%M:%S %p",tz = Sys.timezone())

# create Year, Months, Time, Trip_duration and Hours columns
FHV_list[[10]]$Year = format(FHV_list[[10]]$pickup_datetime, format = "%Y")

FHV_list[[10]]$Months = format(FHV_list[[10]]$pickup_datetime, format = "%m")

FHV_list[[10]]$Time = format(FHV_list[[10]]$pickup_datetime, format = "%H:%M:%S")

FHV_list[[10]]$Trip_duration = difftime(FHV_list[[10]]$dropoff_datetime, FHV_list[[10]]$pickup_datetime, units = "hours")

FHV_list[[10]]$Hours = format(FHV_list[[10]]$pickup_datetime, format = "%H")

FHV_list[[10]]$Weekdays = weekdays(FHV_list[[10]]$pickup_datetime)

FHV_list[[11]]$pickup_datetime = as.POSIXct(FHV_list[[11]]$pickup_datetime, format = "%m/%d/%Y %I:%M:%S %p",tz = Sys.timezone())

FHV_list[[11]]$dropoff_datetime = as.POSIXct(FHV_list[[11]]$dropoff_datetime, format = "%m/%d/%Y %I:%M:%S %p",tz = Sys.timezone())

# create Year, Months, Time, Trip_duration and Hours columns
FHV_list[[11]]$Year = format(FHV_list[[11]]$pickup_datetime, format = "%Y")

FHV_list[[11]]$Months = format(FHV_list[[11]]$pickup_datetime, format = "%m")

FHV_list[[11]]$Time = format(FHV_list[[11]]$pickup_datetime, format = "%H:%M:%S")

FHV_list[[11]]$Trip_duration = difftime(FHV_list[[11]]$dropoff_datetime, FHV_list[[11]]$pickup_datetime, units = "hours")

FHV_list[[11]]$Hours = format(FHV_list[[11]]$pickup_datetime, format = "%H")

FHV_list[[11]]$Weekdays = weekdays(FHV_list[[11]]$pickup_datetime)

FHV_list[[12]]$pickup_datetime = as.POSIXct(FHV_list[[12]]$pickup_datetime, format = "%m/%d/%Y %I:%M:%S %p",tz = Sys.timezone())

FHV_list[[12]]$dropoff_datetime = as.POSIXct(FHV_list[[12]]$dropoff_datetime, format = "%m/%d/%Y %I:%M:%S %p",tz = Sys.timezone())

# create Year, Months, Time, Trip_duration and Hours columns
FHV_list[[12]]$Year = format(FHV_list[[12]]$pickup_datetime, format = "%Y")

FHV_list[[12]]$Months = format(FHV_list[[12]]$pickup_datetime, format = "%m")

FHV_list[[12]]$Time = format(FHV_list[[12]]$pickup_datetime, format = "%H:%M:%S")

FHV_list[[12]]$Trip_duration = difftime(FHV_list[[12]]$dropoff_datetime, FHV_list[[12]]$pickup_datetime, units = "hours")

FHV_list[[12]]$Hours = format(FHV_list[[12]]$pickup_datetime, format = "%H")

FHV_list[[12]]$Weekdays = weekdays(FHV_list[[12]]$pickup_datetime)

FHV_list[[13]]$pickup_datetime = as.POSIXct(FHV_list[[13]]$pickup_datetime, format = "%m/%d/%Y %I:%M:%S %p",tz = Sys.timezone())

FHV_list[[13]]$dropoff_datetime = as.POSIXct(FHV_list[[13]]$dropoff_datetime, format = "%m/%d/%Y %I:%M:%S %p",tz = Sys.timezone())

# create Year, Months, Time, Trip_duration and Hours columns
FHV_list[[13]]$Year = format(FHV_list[[13]]$pickup_datetime, format = "%Y")

FHV_list[[13]]$Months = format(FHV_list[[13]]$pickup_datetime, format = "%m")

FHV_list[[13]]$Time = format(FHV_list[[13]]$pickup_datetime, format = "%H:%M:%S")

FHV_list[[13]]$Trip_duration = difftime(FHV_list[[13]]$dropoff_datetime, FHV_list[[13]]$pickup_datetime, units = "hours")

FHV_list[[13]]$Hours = format(FHV_list[[13]]$pickup_datetime, format = "%H")

FHV_list[[13]]$Weekdays = weekdays(FHV_list[[13]]$pickup_datetime)

FHV_list[[14]]$pickup_datetime = as.POSIXct(FHV_list[[14]]$pickup_datetime, format = "%m/%d/%Y %I:%M:%S %p",tz = Sys.timezone())

FHV_list[[14]]$dropoff_datetime = as.POSIXct(FHV_list[[14]]$dropoff_datetime, format = "%m/%d/%Y %I:%M:%S %p",tz = Sys.timezone())

# create Year, Months, Time, Trip_duration and Hours columns
FHV_list[[14]]$Year = format(FHV_list[[14]]$pickup_datetime, format = "%Y")

FHV_list[[14]]$Months = format(FHV_list[[14]]$pickup_datetime, format = "%m")

FHV_list[[14]]$Time = format(FHV_list[[14]]$pickup_datetime, format = "%H:%M:%S")

FHV_list[[14]]$Trip_duration = difftime(FHV_list[[14]]$dropoff_datetime, FHV_list[[14]]$pickup_datetime, units = "hours")

FHV_list[[14]]$Hours = format(FHV_list[[14]]$pickup_datetime, format = "%H")

FHV_list[[14]]$Weekdays = weekdays(FHV_list[[14]]$pickup_datetime)

# modify global options
options(scipen = 999) 

#write.csv(FHV0, "FHV0.csv")

Calculate sum of trip duration and number of trips of each subset

FHV_group_list = vector(mode = "list",length = 14)

for (i in 1:14){
  FHV_group_list[[i]] = FHV_list[[i]] %>%
  group_by(Year,Months,Hours,Weekdays,dispatching_base_num) %>%
  drop_na(Trip_duration) %>%
  summarize(Total_trip_duration = sum(Trip_duration),
            Total_number_of_trips = n())
}

Combine all the subsets

FHV_group = do.call("rbind", list(FHV_group_list[[1]],FHV_group_list[[2]],FHV_group_list[[3]],
     FHV_group_list[[4]],FHV_group_list[[5]],FHV_group_list[[6]],                               FHV_group_list[[7]],FHV_group_list[[8]],FHV_group_list[[9]],
     FHV_group_list[[10]],FHV_group_list[[11]],FHV_group_list[[12]],
     FHV_group_list[[13]],FHV_group_list[[14]],))

#write.csv(FHV_group,"FHV_group.csv")

FHV_group = read.csv("FHV_group.csv")

A Listing of TLC Licensed Bases

find_a_ride = read.csv("find_a_ride1.csv")
# trim whitespace of License_number
find_a_ride$License_number = trimws(find_a_ride$License_number)

FHV_ride = left_join(FHV_group,find_a_ride, by=(c("dispatching_base_num" = "License_number")))

# extra WAV_Dispatcher from Name_of_Base
FHV_ride$Name_of_Base = str_replace_all(FHV_ride$Name_of_Base, "-" , " ")
FHV_ride$WAV_Dispatcher = word(FHV_ride$Name_of_Base, 1)
FHV_ride$WAV_Dispatcher = str_replace_all(FHV_ride$WAV_Dispatcher, c("FLATIRON" = "Via","GREENPOINT" = "Via"))

# remove "hours" in Total_trip_duration
FHV_ride$Total_trip_duration = as.numeric(word(FHV_ride$Total_trip_duration,1))

# filter out data of Via
FHV_ride = FHV_ride[which(FHV_ride$WAV_Dispatcher != "Via"),]

# convert Months, Hours, Weekdays to factors
FHV_ride[,c(2:5)][]= lapply(FHV_ride[,c(2:5)],as.factor)

#write.csv(FHV_ride[-1],"FHV_ride.csv")

FHV_ride = read.csv("FHV_ride.csv")
FHV_ride

# Total Distance, Total Revenue and Total Trip Duration of NYC Green Taxi in 2020
ggplot(FHV_ride) +
  aes(x = Months, y =  Total_number_of_trips, fill = WAV_Dispatcher) +
  geom_col() + 
  scale_x_continuous(breaks = 1:12) +
  scale_y_continuous(labels = format_format(big.mark = " ", 
                                            decimal.mark = ",", 
                                            scientific = FALSE)) +
 # geom_text(aes(label = Total_number_of_trips), vjust = -1) +
  labs(x = 'Months',
       y = "Number of Trips",
       title = "Number of Trips of NYC 
       Uber and Lyft in 2020") 

 

#  Total Trip Duration in 24 hours
ggplot(FHV_ride) +
  aes(x = Hours, y =  Total_trip_duration, fill = WAV_Dispatcher) +
  geom_col()+
  scale_x_continuous(breaks = 1:23) +
  scale_y_continuous(labels = format_format(big.mark = " ", 
                                            decimal.mark = ",", 
                                            scientific = FALSE)) +
#  geom_text(aes(label = Total_number_of_trips), vjust = -1)+
#  coord_flip()+
  labs(x = 'Hours',
       y = "Total Trip Duration (in hours)",
       title = "Total Trip Duration in 24 hours 
       of Uber and Lyft in 2020") 

ggplot(FHV_ride) +
    aes(x = reorder(Weekdays,Total_trip_duration), y = Total_trip_duration, fill = WAV_Dispatcher) +
    geom_col() +
    scale_y_continuous(labels = format_format(big.mark = " ", 
                                            decimal.mark = ",", 
                                            scientific = FALSE)) +
    labs(x = 'Weekdays',
       y = "Total Trip Duration (in hours)",
       title = "Total Trip Duration in Weekdays 
       of NYC Uber and Lyft in 2020") 

FHV6 = read.csv("FHV6.csv")
FHV6

Calculate trip duration base on location ID of each subset

FHV_location6 = FHV6 %>%
  group_by(PULocationID, DOLocationID,) %>%
  summarize(Total_trip_duration = sum(Trip_duration),
            Total_trips = n()) 

Combine all the FHV_location subsets

FHV_location = rbind(FHV_location[,c(2:5)],FHV_location6)

#write.csv(FHV_location, "FHV_location.csv")

Merge with datasets taxi+_zone_lookup.csv and taxi zone map

Taxi Zone Lookup Table

taxi_zones = read.csv("taxi+_zone_lookup.csv")

Calculate the trip duration and total trips base on pickup location

pickUp = FHV_location %>%
  filter(PULocationID != 264 & PULocationID != 265 & !is.na(Total_trip_duration)) %>% # filter out unknown borough
  group_by(PULocationID) %>%
  summarise(Total_trip_duration = sum(Total_trip_duration),
            Total_trips = sum(Total_trips)) 

Combine pickUp and taxi_zones

pickUp = left_join(pickUp,taxi_zones, by = c("PULocationID" = "LocationID"))

# read GeoJSON file of NYC Taxi zone
js = FROM_GeoJson(url_file_string = "https://data.cityofnewyork.us/api/geospatial/d3c5-ddgc?method=export&format=GeoJSON")

# create a datafram with longitude and latitude of NYC neighborhood 
NYC_coord = lapply(1:length(js$features), 
         function(i){
           if(!is.list(js$features[[i]]$geometry$coordinates)){
           tmpdata <- js$features[[i]]$geometry$coordinates
         }else{
           tmpdata <- js$features[[i]]$geometry$coordinates[[1]]} 
           
tmpdata %>%
   data.frame() %>% 
   tibble %>%
   mutate(zone = js$features[[i]]$properties$zone,
          borough = js$features[[i]]$properties$borough) %>% 
          rename("long" = 'X1', 'lat'='X2')}) %>%
   bind_rows()

pickUp_coord = left_join(pickUp, NYC_coord, by = c("Zone"="zone"))

 

#write.csv(pickUp_coord,"pickUp_coord.csv")

pickUp_coord = read.csv("pickUp_coord.csv")

Top 5 locations

Top_LocationID = pickUp_coord %>%
  filter(Total_trip_duration > 480000) %>%
  count(Zone)
  

Calculate the coordinates of the top 5 locations

center_coord = NYC_coord %>%
  group_by(zone) %>%
  summarise(long = mean(long),lat = mean(lat)) %>%
  right_join(Top_LocationID, by = c("zone" = "Zone"))

#write.csv(center_coord,"center_coord.csv")

ggplot(pickUp_coord)+
  aes(x = long, y = lat,group = Zone, fill =Total_trip_duration) +
  geom_polygon() +
 with(center_coord, 
      annotate(geom="text", x = long, y=lat, 
               label = zone, 
               face="bold",
               size = 5)) +
  scale_fill_gradient(low = "white", high = "blue") +
  labs(title = "Total Trip Duration Hours in Different Pick Up Locations 
       (Uber and Lyft, 2020)")
  

dropOff = FHV_location %>%
  filter(DOLocationID != 264 & DOLocationID != 265 & !is.na(Total_trip_duration)) %>% # filter out unknown borough
  group_by(DOLocationID) %>%
  summarise(Total_trip_duration = sum(Total_trip_duration),
            Total_trips = sum(Total_trips)) 
dropOff

dropOff_coord = left_join(dropOff, NYC_coord, by = c("Zone"="zone"))

Top_DLocationID = dropOff_coord %>%
  filter(!is.na(Zone)) %>%
  filter(Total_trip_duration > 390000) %>%
  count(Zone)
  
Top_DLocationID  

center_Dcoord = NYC_coord %>%
  group_by(zone) %>%
  summarise(long = mean(long),lat = mean(lat)) %>%
  right_join(Top_DLocationID, by = c("zone" = "Zone"))
center_Dcoord

#write.csv(center_Dcoord,"center_Dcoord.csv")

ggplot(dropOff_coord)+
  aes(x = long, y = lat,group = Zone, fill =Total_trips) +
  geom_polygon() +
  with(center_Dcoord, 
      annotate(geom="text", x = long, y=lat, 
               label = zone,
               face="bold",
               size = 5)) +
  scale_fill_gradient(low = "white", high = "red") +
  labs(title = "Total Trips in Different Pick Up Locations 
       (Uber and Lyft, 2020)")