Exercise 05: Text Data Analysis and Data Summary

Objective

Use the nycflights13 dataset to analyze flight delays and apply key dplyr functions.

Instructions

Using the flights dataset from the nycflights13 package:

1. Filter flights that departed from JFK in June.
   • Hint origin == "JFK" and month == 6
2. Select relevant columns: year, month, day, dep_time, arr_time, carrier, flight, and dep_delay.
3. Arrange flights in descending order of departure delay.
4. Create a new column called delay_category:
   • "On Time" if dep_delay <= 0
   • "Slight Delay" if dep_delay is between 1 and 30 minutes
   • "Severe Delay" if dep_delay is greater than 30 minutes
5. Summarize the average delay for each airline (carrier).

Questions

1. What proportion of flights from JFK in June were categorized as “Severe Delay”?
2. Which carrier had the second longest average departure delay?

Answer

# Load required packages
library(nycflights13) # install.packages("nycflights13")
library(tidyverse)

# Step 1: Filter flights departing from JFK in June
jfk_june_flights <- flights |> 
  filter(origin == "JFK", month == 6)

# Step 2: Select relevant columns
jfk_june_flights_selected <- jfk_june_flights |> 
  select(year, month, day, dep_time, arr_time, carrier, flight, dep_delay)

# Step 3: Arrange by departure delay (descending order)
jfk_june_flights_sorted <- jfk_june_flights_selected |> 
  arrange(desc(dep_delay))

# Step 4: Create delay categories
jfk_june_flights_categorized <- jfk_june_flights_sorted |> 
  mutate(
    delay_category = case_when(
      dep_delay <= 0 ~ "On Time",
      dep_delay > 0 & dep_delay <= 30 ~ "Slight Delay",
      dep_delay > 30 ~ "Severe Delay",
      is.na(dep_delay) ~ "Dont Know" # character type value
    )
  )
mean(jfk_june_flights_categorized$delay_category == "Severe Delay") # 20.1%
mean(c(TRUE, FALSE, FALSE))

# Step 5: Summarize average delay by carrier
avg_delay_by_carrier <- jfk_june_flights_categorized |> 
  group_by(carrier) |> 
  summarize(avg_dep_delay = mean(dep_delay, na.rm = TRUE)) |> 
  arrange(desc(avg_dep_delay))
avg_delay_by_carrier # Q1 Answer: VX carrier: Virgin America.

# Step 6: Visualize the average delay per airline
ggplot(avg_delay_by_carrier, aes(x = fct_reorder(carrier, avg_dep_delay), y = avg_dep_delay)) +
  geom_col(fill = "steelblue") +
  geom_label(aes(label = round(avg_dep_delay, 1))) +
  labs(title = "Average Departure Delay by Carrier (JFK, June)",
       x = "Carrier", y = "Avg Departure Delay (min)") +
  coord_flip()