largest causes of flight delays and cancellations
What are the largest causes of flight delays and cancellations?
This Project will use the nycflights13 data package for all data and visualizations. I also use the tidyverse library and the dplyr for data manipulation and visualizations.
suppressWarnings(library(tidyverse))## ── Attaching core tidyverse packages ──────────────────────── tidyverse 2.0.0 ──
## ✔ dplyr 1.1.2 ✔ readr 2.1.4
## ✔ forcats 1.0.0 ✔ stringr 1.5.0
## ✔ ggplot2 3.4.3 ✔ tibble 3.2.1
## ✔ lubridate 1.9.2 ✔ tidyr 1.3.0
## ✔ purrr 1.0.2
## ── Conflicts ────────────────────────────────────────── tidyverse_conflicts() ──
## ✖ dplyr::filter() masks stats::filter()
## ✖ dplyr::lag() masks stats::lag()
## ℹ Use the conflicted package (<http://conflicted.r-lib.org/>) to force all conflicts to become errorssuppressWarnings(library(nycflights13))
suppressWarnings(library(dplyr))this project will use the nycflights13 dataset for data and visualizations but due to its size and stress on loading the entire dataset, I will limit the amount shown to only the first 10 entries.
slice(flights, 1:10)This dataset has 3367766 observations with 19 variables.
Below is a data set that starts to outline us exploring some of the data’s basic trends. This data shows days that flights were canceled or delayed by more than one hour where more then 35% were canceled or delayed.
cancel_delay_counts <- flights %>%
mutate(is_cancelled_or_delayed = is.na(dep_time) | dep_delay > 60) %>%
group_by(year, month, day) %>%
summarise(
total_flights = n(),
canceled_or_delayed = sum(is_cancelled_or_delayed),
percent_cancelled_or_delayed = (canceled_or_delayed / total_flights) * 100,
.groups = 'drop'
)
high_cancel_delay_days <- cancel_delay_counts %>%
filter(percent_cancelled_or_delayed > 35)
print(high_cancel_delay_days)## # A tibble: 12 × 6
## year month day total_flights canceled_or_delayed percent_cancelled_or_de…¹
## <int> <int> <int> <int> <int> <dbl>
## 1 2013 2 8 930 506 54.4
## 2 2013 2 9 684 419 61.3
## 3 2013 3 8 979 574 58.6
## 4 2013 5 23 988 435 44.0
## 5 2013 6 24 994 348 35.0
## 6 2013 6 28 994 369 37.1
## 7 2013 7 1 966 407 42.1
## 8 2013 7 10 1004 364 36.3
## 9 2013 7 23 997 362 36.3
## 10 2013 9 2 929 326 35.1
## 11 2013 9 12 992 404 40.7
## 12 2013 12 5 969 385 39.7
## # ℹ abbreviated name: ¹percent_cancelled_or_delayedFrom this we can see that a majority of the delays and cancellations take place during winter months there blizzards and snow storms are extremely common and looking at weather reports of these days from 2013 reveal that these days had bad weather and large storms that took place on multiple of these days.
Below is a data set that shows each scheduled departure hour during the day, the percentage of flights that were canceled or delayed by more than one hour.
cancel_delay_counts <- flights %>%
mutate(is_cancelled_or_delayed = is.na(dep_time) | dep_delay > 60) %>%
group_by(hour) %>%
summarise(
total_flights = n(),
canceled_or_delayed = sum(is_cancelled_or_delayed),
percent_cancelled_or_delayed = (canceled_or_delayed / total_flights) * 100) %>%
filter(hour != 1)
high_cancel_delay_days <- flights %>%
mutate(is_cancelled_or_delayed = is.na(dep_time) | dep_delay > 60) %>%
group_by(month, day) %>%
summarise(
total_flights = n(),
canceled_or_delayed = sum(is_cancelled_or_delayed),
percent_cancelled_or_delayed = (canceled_or_delayed / total_flights) * 100,
.groups = 'drop') %>%
filter(percent_cancelled_or_delayed >= 35)
print(high_cancel_delay_days)## # A tibble: 12 × 5
## month day total_flights canceled_or_delayed percent_cancelled_or_delayed
## <int> <int> <int> <int> <dbl>
## 1 2 8 930 506 54.4
## 2 2 9 684 419 61.3
## 3 3 8 979 574 58.6
## 4 5 23 988 435 44.0
## 5 6 24 994 348 35.0
## 6 6 28 994 369 37.1
## 7 7 1 966 407 42.1
## 8 7 10 1004 364 36.3
## 9 7 23 997 362 36.3
## 10 9 2 929 326 35.1
## 11 9 12 992 404 40.7
## 12 12 5 969 385 39.7This data set just continues to show how much this severe weather effected our flights on these days take 2/9 for example 61% of flights canceled is a huge portion of the airports flights per day especially when on average a airport is getting between 1,000-3,00 flights a day.
The worst times to fly and why.
Below is a graph showing the relationship between the time and the percentage of flights canceled or delayed by more then an hour during that time. As time goes on we can clearly see a trend in the graph. The key thing to pay atrention to is how more delays occur as time goes on.
ggplot(cancel_delay_counts, aes(x = hour, y = percent_cancelled_or_delayed)) +
geom_line() +
labs(
title = "Percentage of flights cancelled or delayed",
x = "Hour",
y = "Percent Cancelled or Delayed",
caption = "Data from: nycflights13"
)
To address the outlier right off the bat I have filtered out the major outlier which was a single flight that was canceled at 1am. This flight skewed the data because this outlier gave us a 100% cancellation rate at 1am.
In general delays and cancellations increase as the day does on. This is because of a lot of effects such as airports getting busier and congested. If one plane takes too long it can create a backup leading to more flights being delayed as each just compounds the issue.
Overall the riskiest time to fly is between 8pm to 10pm or 20-22 in military time if you want to avoid cancellations and long delays. The delays at this time could also be when pilots or flight staff change shifts.
The worst days for cancelations and delays
Above you saw see data listing off the significance days where a lot of cancellations and delays took place. However, some planes did manage to leave on these days and below you’ll see the planes that made it out.
filtered_flights <- flights %>%
filter(
(month == 2 & day == 8) |
(month == 2 & day == 9) |
(month == 3 & day == 8) |
(month == 5 & day == 23) |
(month == 6 & day == 24) |
(month == 6 & day == 28) |
(month == 7 & day == 1) |
(month == 7 & day == 10) |
(month == 7 & day == 23) |
(month == 9 & day == 12) |
(month == 12 & day == 5) |
(month == 9 & day == 2)
) %>%
filter(dep_delay <= 0)
print(filtered_flights)## # A tibble: 3,441 × 19
## year month day dep_time sched_dep_time dep_delay arr_time sched_arr_time
## <int> <int> <int> <int> <int> <dbl> <int> <int>
## 1 2013 12 5 457 500 -3 637 651
## 2 2013 12 5 512 515 -3 753 814
## 3 2013 12 5 527 530 -3 657 706
## 4 2013 12 5 539 540 -1 832 850
## 5 2013 12 5 540 545 -5 822 832
## 6 2013 12 5 544 550 -6 959 1027
## 7 2013 12 5 548 600 -12 738 755
## 8 2013 12 5 551 600 -9 804 810
## 9 2013 12 5 553 600 -7 919 915
## 10 2013 12 5 553 600 -7 645 701
## # ℹ 3,431 more rows
## # ℹ 11 more variables: arr_delay <dbl>, carrier <chr>, flight <int>,
## # tailnum <chr>, origin <chr>, dest <chr>, air_time <dbl>, distance <dbl>,
## # hour <dbl>, minute <dbl>, time_hour <dttm>With this data you’ll see that a majority of the planes that made it out on time or early left early in the morning before the worst of the weather or storms could hit.
Above we could see that the planes that were able to get out on time seem to be the one that leave early but below breaking down the data to see average time planes left will help show average times.
average_actual_hour <- flights %>%
filter(
(month == 2 & day == 8) |
(month == 2 & day == 9) |
(month == 3 & day == 8) |
(month == 5 & day == 23) |
(month == 6 & day == 24) |
(month == 6 & day == 28) |
(month == 7 & day == 1) |
(month == 7 & day == 10) |
(month == 7 & day == 23) |
(month == 9 & day == 12) |
(month == 9 & day == 2) |
(month == 12 & day == 5)
) %>%
filter(!is.na(dep_time)) %>%
mutate(
dep_hour = dep_time %/% 100,
dep_min = dep_time %% 100,
dep_time_decimal = dep_hour + (dep_min / 60)
) %>%
group_by(month, day) %>%
summarise(average_actual_hour = mean(dep_time_decimal, na.rm = TRUE),
.groups = 'drop')
print(average_actual_hour)## # A tibble: 12 × 3
## month day average_actual_hour
## <int> <int> <dbl>
## 1 2 8 10.1
## 2 2 9 17.1
## 3 3 8 14.4
## 4 5 23 13.1
## 5 6 24 13.3
## 6 6 28 13.2
## 7 7 1 13.8
## 8 7 10 13.2
## 9 7 23 14.1
## 10 9 2 14.1
## 11 9 12 12.7
## 12 12 5 14.0This table shows the average hour planes on the days with a lot of delays that left. Most left fairly early and well before dark and before the storms, which it seems is the most likely reason for why most plans were able to leave. But what about the day 2/9? Why is the average so much higher then the others? It seems like that day around that time there was a break in the snowy weather with the wind speeds dropping significantly. This seems like the reason planes were able to leave. Other days in the afternoon seem to correlate with the temperature rising around 12-2pm.
In summary
Overall the weather especially winter weather has an extreme effect on air traffic and whether its delayed or not and these issues are only increased when we add other effects such as flights getting backed up as the day goes on. This data helps prove that some of the worst times attempt to have a flight leave is late in the day after dark and with inclement weather such as big thunder storms or blizzards. We also see that flights between 8pm to 10pm are some of the worst to have if your plan is to leave on time. As the day goes on flights back up and these back ups continue to pile up leading to large amounts of delays around 8pm to 10pm but after these times the airlines get caught back up. So leaving on days with good weather and clear skies between 5am-10am will give you the best chance of leaving on time.