Time Series Analysis:
Relationship Between Temperature and Weather Flight Delays in Boston Area (2006-08 to 2016-08)
Relationship Between Temperature and Weather Flight Delays in Boston Area (2006-08 to 2016-08)
By Danny
Introuduction
This page shows the non-linear relationship between the temperature and the Weather Flight Delays by plotting Time Series Plot.
Install and Load the packages
library(readr)
suppressMessages(library(zoo))
suppressMessages(library(dplyr))
suppressMessages(library(xts))
suppressMessages(library(plotly))Loading the dataset : Boston_flight_information from local directory, and name it flights
BOS_airport <- read_csv("C:/Users/User/Desktop/BOS_airport.csv")## Parsed with column specification:
## cols(
## .default = col_integer(),
## carrier = col_character(),
## carrier_name = col_character(),
## airport = col_character(),
## airport_name = col_character(),
## carrier_ct = col_double(),
## weather_ct = col_double(),
## nas_ct = col_double(),
## security_ct = col_double(),
## late_aircraft_ct = col_double()
## )## See spec(...) for full column specifications.flights <- BOS_airport
head(flights,10)## # A tibble: 10 × 21
## year month carrier carrier_name airport
## <int> <int> <chr> <chr> <chr>
## 1 2003 6 AA American Airlines Inc. BOS
## 2 2003 6 AS Alaska Airlines Inc. BOS
## 3 2003 6 CO Continental Air Lines Inc. BOS
## 4 2003 6 DH Atlantic Coast Airlines BOS
## 5 2003 6 DL Delta Air Lines Inc. BOS
## 6 2003 6 EV Atlantic Southeast Airlines BOS
## 7 2003 6 FL AirTran Airways Corporation BOS
## 8 2003 6 HP America West Airlines Inc. BOS
## 9 2003 6 MQ American Eagle Airlines Inc. BOS
## 10 2003 6 NW Northwest Airlines Inc. BOS
## # ... with 16 more variables: airport_name <chr>, arr_flights <int>,
## # arr_del15 <int>, carrier_ct <dbl>, weather_ct <dbl>, nas_ct <dbl>,
## # security_ct <dbl>, late_aircraft_ct <dbl>, arr_cancelled <int>,
## # arr_diverted <int>, arr_delay <int>, carrier_delay <int>,
## # weather_delay <int>, nas_delay <int>, security_delay <int>,
## # late_aircraft_delay <int>Covnert date into acceptable format for “as.Date fucntion”
flights$date <- paste(flights$year,sprintf("%02d",flights$month), sep= "-")
flights$days <- c(1)
flights$date <- paste(flights$date,sprintf("%02d",flights$days), sep= "-")
flights$date<- as.Date(flights$date)
flights<-flights %>% select(date, everything())
head(flights,10) ## multiple same date is because there are different airline companys recorded## # A tibble: 10 × 23
## date year month carrier carrier_name airport
## <date> <int> <int> <chr> <chr> <chr>
## 1 2003-06-01 2003 6 AA American Airlines Inc. BOS
## 2 2003-06-01 2003 6 AS Alaska Airlines Inc. BOS
## 3 2003-06-01 2003 6 CO Continental Air Lines Inc. BOS
## 4 2003-06-01 2003 6 DH Atlantic Coast Airlines BOS
## 5 2003-06-01 2003 6 DL Delta Air Lines Inc. BOS
## 6 2003-06-01 2003 6 EV Atlantic Southeast Airlines BOS
## 7 2003-06-01 2003 6 FL AirTran Airways Corporation BOS
## 8 2003-06-01 2003 6 HP America West Airlines Inc. BOS
## 9 2003-06-01 2003 6 MQ American Eagle Airlines Inc. BOS
## 10 2003-06-01 2003 6 NW Northwest Airlines Inc. BOS
## # ... with 17 more variables: airport_name <chr>, arr_flights <int>,
## # arr_del15 <int>, carrier_ct <dbl>, weather_ct <dbl>, nas_ct <dbl>,
## # security_ct <dbl>, late_aircraft_ct <dbl>, arr_cancelled <int>,
## # arr_diverted <int>, arr_delay <int>, carrier_delay <int>,
## # weather_delay <int>, nas_delay <int>, security_delay <int>,
## # late_aircraft_delay <int>, days <dbl>range(flights$date)## [1] "2003-06-01" "2016-09-01"Select Weather_delayed column and date, then convert both columns into Xts format
Weather Flight Delays column indicates the delay caused by the weather
weather_flights_delays <- select(flights , c(1,19))
weather_flights_delays_xts<- as.xts(weather_flights_delays[,-1],order.by = weather_flights_delays$date)
head(weather_flights_delays_xts,10)## weather_delay
## 2003-06-01 1783
## 2003-06-01 0
## 2003-06-01 194
## 2003-06-01 312
## 2003-06-01 275
## 2003-06-01 0
## 2003-06-01 25
## 2003-06-01 0
## 2003-06-01 317
## 2003-06-01 550Sum the the number of Weather_delay flights by month
weather_flights_delays_xts<- apply.monthly(weather_flights_delays_xts, sum)
colnames(weather_flights_delays_xts) <- c("weather_flights_delays")
head(weather_flights_delays_xts,10)## weather_flights_delays
## 2003-06-01 4160
## 2003-07-01 5024
## 2003-08-01 7368
## 2003-09-01 1751
## 2003-10-01 653
## 2003-11-01 1514
## 2003-12-01 4979
## 2004-01-01 4580
## 2004-02-01 3425
## 2004-03-01 2249Fetch Boston temperature data from URL
library("weatherData")
weatherData2006_2016 <- do.call(
rbind,
lapply(
2006:2016,
function(i) getWeatherForYear(station_id = "BOS", year = i)
)
)## [1] "CET" "Max_TemperatureC"
## [3] "Mean_TemperatureC" "Min_TemperatureC"
## [5] "Dew_PointC" "MeanDew_PointC"
## [7] "Min_DewpointC" "Max_Humidity"
## [9] "Mean_Humidity" "Min_Humidity"
## [11] "Max_Sea_Level_PressurehPa" "Mean_Sea_Level_PressurehPa"
## [13] "Min_Sea_Level_PressurehPa" "Max_VisibilityKm"
## [15] "Mean_VisibilityKm" "Min_VisibilitykM"
## [17] "Max_Wind_SpeedKm_h" "Mean_Wind_SpeedKm_h"
## [19] "Max_Gust_SpeedKm_h" "Precipitationmm"
## [21] "CloudCover" "Events"
## [23] "WindDirDegrees"
## [1] "Date" "Max_TemperatureC" "Mean_TemperatureC"
## [4] "Min_TemperatureC"
## [1] "CET" "Max_TemperatureC"
## [3] "Mean_TemperatureC" "Min_TemperatureC"
## [5] "Dew_PointC" "MeanDew_PointC"
## [7] "Min_DewpointC" "Max_Humidity"
## [9] "Mean_Humidity" "Min_Humidity"
## [11] "Max_Sea_Level_PressurehPa" "Mean_Sea_Level_PressurehPa"
## [13] "Min_Sea_Level_PressurehPa" "Max_VisibilityKm"
## [15] "Mean_VisibilityKm" "Min_VisibilitykM"
## [17] "Max_Wind_SpeedKm_h" "Mean_Wind_SpeedKm_h"
## [19] "Max_Gust_SpeedKm_h" "Precipitationmm"
## [21] "CloudCover" "Events"
## [23] "WindDirDegrees"
## [1] "Date" "Max_TemperatureC" "Mean_TemperatureC"
## [4] "Min_TemperatureC"
## [1] "CET" "Max_TemperatureC"
## [3] "Mean_TemperatureC" "Min_TemperatureC"
## [5] "Dew_PointC" "MeanDew_PointC"
## [7] "Min_DewpointC" "Max_Humidity"
## [9] "Mean_Humidity" "Min_Humidity"
## [11] "Max_Sea_Level_PressurehPa" "Mean_Sea_Level_PressurehPa"
## [13] "Min_Sea_Level_PressurehPa" "Max_VisibilityKm"
## [15] "Mean_VisibilityKm" "Min_VisibilitykM"
## [17] "Max_Wind_SpeedKm_h" "Mean_Wind_SpeedKm_h"
## [19] "Max_Gust_SpeedKm_h" "Precipitationmm"
## [21] "CloudCover" "Events"
## [23] "WindDirDegrees"
## [1] "Date" "Max_TemperatureC" "Mean_TemperatureC"
## [4] "Min_TemperatureC"
## [1] "CET" "Max_TemperatureC"
## [3] "Mean_TemperatureC" "Min_TemperatureC"
## [5] "Dew_PointC" "MeanDew_PointC"
## [7] "Min_DewpointC" "Max_Humidity"
## [9] "Mean_Humidity" "Min_Humidity"
## [11] "Max_Sea_Level_PressurehPa" "Mean_Sea_Level_PressurehPa"
## [13] "Min_Sea_Level_PressurehPa" "Max_VisibilityKm"
## [15] "Mean_VisibilityKm" "Min_VisibilitykM"
## [17] "Max_Wind_SpeedKm_h" "Mean_Wind_SpeedKm_h"
## [19] "Max_Gust_SpeedKm_h" "Precipitationmm"
## [21] "CloudCover" "Events"
## [23] "WindDirDegrees"
## [1] "Date" "Max_TemperatureC" "Mean_TemperatureC"
## [4] "Min_TemperatureC"
## [1] "CET" "Max_TemperatureC"
## [3] "Mean_TemperatureC" "Min_TemperatureC"
## [5] "Dew_PointC" "MeanDew_PointC"
## [7] "Min_DewpointC" "Max_Humidity"
## [9] "Mean_Humidity" "Min_Humidity"
## [11] "Max_Sea_Level_PressurehPa" "Mean_Sea_Level_PressurehPa"
## [13] "Min_Sea_Level_PressurehPa" "Max_VisibilityKm"
## [15] "Mean_VisibilityKm" "Min_VisibilitykM"
## [17] "Max_Wind_SpeedKm_h" "Mean_Wind_SpeedKm_h"
## [19] "Max_Gust_SpeedKm_h" "Precipitationmm"
## [21] "CloudCover" "Events"
## [23] "WindDirDegrees"
## [1] "Date" "Max_TemperatureC" "Mean_TemperatureC"
## [4] "Min_TemperatureC"
## [1] "CET" "Max_TemperatureC"
## [3] "Mean_TemperatureC" "Min_TemperatureC"
## [5] "Dew_PointC" "MeanDew_PointC"
## [7] "Min_DewpointC" "Max_Humidity"
## [9] "Mean_Humidity" "Min_Humidity"
## [11] "Max_Sea_Level_PressurehPa" "Mean_Sea_Level_PressurehPa"
## [13] "Min_Sea_Level_PressurehPa" "Max_VisibilityKm"
## [15] "Mean_VisibilityKm" "Min_VisibilitykM"
## [17] "Max_Wind_SpeedKm_h" "Mean_Wind_SpeedKm_h"
## [19] "Max_Gust_SpeedKm_h" "Precipitationmm"
## [21] "CloudCover" "Events"
## [23] "WindDirDegrees"
## [1] "Date" "Max_TemperatureC" "Mean_TemperatureC"
## [4] "Min_TemperatureC"
## [1] "CET" "Max_TemperatureC"
## [3] "Mean_TemperatureC" "Min_TemperatureC"
## [5] "Dew_PointC" "MeanDew_PointC"
## [7] "Min_DewpointC" "Max_Humidity"
## [9] "Mean_Humidity" "Min_Humidity"
## [11] "Max_Sea_Level_PressurehPa" "Mean_Sea_Level_PressurehPa"
## [13] "Min_Sea_Level_PressurehPa" "Max_VisibilityKm"
## [15] "Mean_VisibilityKm" "Min_VisibilitykM"
## [17] "Max_Wind_SpeedKm_h" "Mean_Wind_SpeedKm_h"
## [19] "Max_Gust_SpeedKm_h" "Precipitationmm"
## [21] "CloudCover" "Events"
## [23] "WindDirDegrees"
## [1] "Date" "Max_TemperatureC" "Mean_TemperatureC"
## [4] "Min_TemperatureC"
## [1] "CET" "Max_TemperatureC"
## [3] "Mean_TemperatureC" "Min_TemperatureC"
## [5] "Dew_PointC" "MeanDew_PointC"
## [7] "Min_DewpointC" "Max_Humidity"
## [9] "Mean_Humidity" "Min_Humidity"
## [11] "Max_Sea_Level_PressurehPa" "Mean_Sea_Level_PressurehPa"
## [13] "Min_Sea_Level_PressurehPa" "Max_VisibilityKm"
## [15] "Mean_VisibilityKm" "Min_VisibilitykM"
## [17] "Max_Wind_SpeedKm_h" "Mean_Wind_SpeedKm_h"
## [19] "Max_Gust_SpeedKm_h" "Precipitationmm"
## [21] "CloudCover" "Events"
## [23] "WindDirDegrees"
## [1] "Date" "Max_TemperatureC" "Mean_TemperatureC"
## [4] "Min_TemperatureC"
## [1] "CET" "Max_TemperatureC"
## [3] "Mean_TemperatureC" "Min_TemperatureC"
## [5] "Dew_PointC" "MeanDew_PointC"
## [7] "Min_DewpointC" "Max_Humidity"
## [9] "Mean_Humidity" "Min_Humidity"
## [11] "Max_Sea_Level_PressurehPa" "Mean_Sea_Level_PressurehPa"
## [13] "Min_Sea_Level_PressurehPa" "Max_VisibilityKm"
## [15] "Mean_VisibilityKm" "Min_VisibilitykM"
## [17] "Max_Wind_SpeedKm_h" "Mean_Wind_SpeedKm_h"
## [19] "Max_Gust_SpeedKm_h" "Precipitationmm"
## [21] "CloudCover" "Events"
## [23] "WindDirDegrees"
## [1] "Date" "Max_TemperatureC" "Mean_TemperatureC"
## [4] "Min_TemperatureC"
## [1] "CET" "Max_TemperatureC"
## [3] "Mean_TemperatureC" "Min_TemperatureC"
## [5] "Dew_PointC" "MeanDew_PointC"
## [7] "Min_DewpointC" "Max_Humidity"
## [9] "Mean_Humidity" "Min_Humidity"
## [11] "Max_Sea_Level_PressurehPa" "Mean_Sea_Level_PressurehPa"
## [13] "Min_Sea_Level_PressurehPa" "Max_VisibilityKm"
## [15] "Mean_VisibilityKm" "Min_VisibilitykM"
## [17] "Max_Wind_SpeedKm_h" "Mean_Wind_SpeedKm_h"
## [19] "Max_Gust_SpeedKm_h" "Precipitationmm"
## [21] "CloudCover" "Events"
## [23] "WindDirDegrees"
## [1] "Date" "Max_TemperatureC" "Mean_TemperatureC"
## [4] "Min_TemperatureC"
## [1] "CET" "Max_TemperatureC"
## [3] "Mean_TemperatureC" "Min_TemperatureC"
## [5] "Dew_PointC" "MeanDew_PointC"
## [7] "Min_DewpointC" "Max_Humidity"
## [9] "Mean_Humidity" "Min_Humidity"
## [11] "Max_Sea_Level_PressurehPa" "Mean_Sea_Level_PressurehPa"
## [13] "Min_Sea_Level_PressurehPa" "Max_VisibilityKm"
## [15] "Mean_VisibilityKm" "Min_VisibilitykM"
## [17] "Max_Wind_SpeedKm_h" "Mean_Wind_SpeedKm_h"
## [19] "Max_Gust_SpeedKm_h" "Precipitationmm"
## [21] "CloudCover" "Events"
## [23] "WindDirDegrees"
## [1] "Date" "Max_TemperatureC" "Mean_TemperatureC"
## [4] "Min_TemperatureC"head(weatherData2006_2016,10)## Date Max_TemperatureC Mean_TemperatureC Min_TemperatureC
## 1 2006-01-01 8 6 3
## 2 2006-01-02 8 4 2
## 3 2006-01-03 8 4 2
## 4 2006-01-04 5 3 2
## 5 2006-01-05 4 2 0
## 6 2006-01-06 6 3 1
## 7 2006-01-07 6 3 1
## 8 2006-01-08 3 1 -1
## 9 2006-01-09 3 0 -3
## 10 2006-01-10 2 -1 -4Change all the dates in order to being concerted with weather_delay_flights_xts
weatherData2006_2016$Date<- substring(as.character(weatherData2006_2016$Date), 1,8)
weatherData2006_2016$days <- c(1)
weatherData2006_2016$Date <- paste(weatherData2006_2016$Date,sprintf("%02d",weatherData2006_2016$days), sep= "")
weatherData2006_2016$Date <- as.Date(weatherData2006_2016$Date)
weatherData2006_2016$Mean_TemperatureC<-na.locf(weatherData2006_2016$Mean_TemperatureC)
summary(weatherData2006_2016)## Date Max_TemperatureC Mean_TemperatureC Min_TemperatureC
## Min. :2006-01-01 Min. :-7.00 Min. :-9.00 Min. :-15.000
## 1st Qu.:2008-10-01 1st Qu.:10.00 1st Qu.: 6.00 1st Qu.: 1.000
## Median :2011-07-01 Median :17.00 Median :11.00 Median : 6.000
## Mean :2011-06-17 Mean :16.94 Mean :11.36 Mean : 6.097
## 3rd Qu.:2014-04-01 3rd Qu.:24.00 3rd Qu.:18.00 3rd Qu.: 12.000
## Max. :2016-12-01 Max. :37.00 Max. :28.00 Max. : 25.000
## days
## Min. :1
## 1st Qu.:1
## Median :1
## Mean :1
## 3rd Qu.:1
## Max. :1 weatherData2006_2016<-na.omit(weatherData2006_2016)
head(weatherData2006_2016$Date,10)## [1] "2006-01-01" "2006-01-01" "2006-01-01" "2006-01-01" "2006-01-01"
## [6] "2006-01-01" "2006-01-01" "2006-01-01" "2006-01-01" "2006-01-01"Convert weatherData2006_2016 to xts format, then calculate the monthly average temperature(2006-08 to 2016-08)
weatherData2006_2016_xts <- xts(weatherData2006_2016[-1], order.by = weatherData2006_2016$Date)
Monthly_Temp_Mean<- apply.monthly(weatherData2006_2016_xts[,2], mean)
head(Monthly_Temp_Mean,10)## Mean_TemperatureC
## 2006-01-01 -0.7741935
## 2006-02-01 1.2142857
## 2006-03-01 5.1935484
## 2006-04-01 10.3000000
## 2006-05-01 13.8387097
## 2006-06-01 17.1666667
## 2006-07-01 19.6129032
## 2006-08-01 19.3548387
## 2006-09-01 16.4000000
## 2006-10-01 12.0967742Plot weather_delay flight and monthly average temperature individually
Total_Number_of_Weather_flight_delays <- weather_flights_delays_xts["2006-08-01/2016-08-01"]
Monthly_Average_Temperature_Boston <- Monthly_Temp_Mean["2006-08-01/2016-08-01"]
plot.xts(Total_Number_of_Weather_flight_delays, major.format = "%y",xlab="Year",ylab="Flights")
plot.xts(Monthly_Average_Temperature_Boston,major.format = "%y",xlab="Year",ylab="Temperature")
Plot weather delay flight and monthly average temperature together
Weather_Flights_Delay_Vs_Temperature <- merge(Total_Number_of_Weather_flight_delays,Monthly_Average_Temperature_Boston)
colnames(Weather_Flights_Delay_Vs_Temperature) <- c("Flights","Temperature")
plot.zoo(Weather_Flights_Delay_Vs_Temperature, plot.type = "multiple", xlab = "Year",col= c("blue", "red"),main = "Weather Flights Delay VS. Temperature in Boston",sub="My Sub-title")