Load the required libraries

library(tidyr)
library(dplyr)
library(ggplot2)
library(flextable)
library(rstatix)

GitHub URL for the CSV file

url <- "https://raw.githubusercontent.com/jewelercart/Data607/main/flight_data.csv"

Read the CSV file into R, specifying header and line termination

flight_data <- read.csv(url, header = TRUE, sep = ",", quote = "", fill = TRUE)

Specify the column names explicitly

colnames(flight_data) <- c("Time_Zone", "Cities", "on_time", "delayed")

Tidy the data (convert to long format)

flight_data_long <- flight_data %>%
  gather(key = "Status", value = "Count", -Time_Zone, -Cities)

Removing the quotes from the names of Time Zones and Cities.

flight_data_long$Time_Zone<- gsub("\"", "", flight_data_long$Time_Zone)
flight_data_long$Cities<- gsub("\"", "", flight_data_long$Cities)
table <- knitr::kable(flight_data_long)
table
Time_Zone Cities Status Count
ALASKA Los Angeles on_time 497
ALASKA Phoenix on_time 221
ALASKA San Diego on_time 212
ALASKA San Francisco on_time 503
ALASKA Seattle on_time 1841
AM WEST Los Angeles on_time 694
AM WEST Phoenix on_time 4840
AM WEST San Diego on_time 383
AM WEST San Francisco on_time 320
AM WEST Seattle on_time 201
ALASKA Los Angeles delayed 62
ALASKA Phoenix delayed 12
ALASKA San Diego delayed 20
ALASKA San Francisco delayed 102
ALASKA Seattle delayed 305
AM WEST Los Angeles delayed 117
AM WEST Phoenix delayed 415
AM WEST San Diego delayed 65
AM WEST San Francisco delayed 129
AM WEST Seattle delayed 61

Calculate summary statistics, handling missing values

summary_stats =
flight_data_long  %>%
  group_by(Time_Zone, Status) %>%
   get_summary_stats(Count, show = c("mean", "median", "max","min"))

Create bar plots to compare arrival delays

ggplot(flight_data_long, aes(x = Time_Zone, y = Count, fill = Status)) +
  geom_bar(stat = "identity", position = "dodge") +
  labs(title = "Comparison of Arrival Delays by Time Zones and Status", y = "Number of Flights") +
  theme_minimal()

Filter the data for ALASKA and AM WEST separately:

alaska_data <- flight_data_long %>% filter(Time_Zone == "ALASKA")
am_west_data <- flight_data_long %>% filter(Time_Zone == "AM WEST")

Create bar plots for ALASKA and AM WEST:

plot_alaska <- ggplot(alaska_data, aes(x = Cities, y = Count, fill = Status)) +
  geom_bar(stat = "identity", position = "dodge") +
  labs(title = "Arrival Delays for ALASKA by City and Status", y = "Number of Flights") +
  theme_minimal() +
  theme(axis.text.x = element_text(angle = 45, hjust = 1),
        plot.title = element_text(hjust=0.5),
        panel.grid.minor= element_blank())

plot_alaska

plot_am_west <- ggplot(am_west_data, aes(x = Cities, y = Count, fill = Status)) +
  geom_bar(stat = "identity", position = "dodge") +
  labs(title = "Arrival Delays for AM WEST by City and Status", y = "Count") +
  theme_minimal() +
  theme(axis.text.x = element_text(angle = 45, hjust = 1),
        plot.title = element_text(hjust=0.5),
        panel.grid.minor= element_blank())

plot_am_west

Summary Statistics:

summary_stats
## # A tibble: 4 × 8
##   Time_Zone Status  variable     n  mean median   max   min
##   <chr>     <chr>   <fct>    <dbl> <dbl>  <dbl> <dbl> <dbl>
## 1 ALASKA    delayed Count        5  100.     62   305    12
## 2 ALASKA    on_time Count        5  655.    497  1841   212
## 3 AM WEST   delayed Count        5  157.    117   415    61
## 4 AM WEST   on_time Count        5 1288.    383  4840   201

A Chi-square test will compare whether the delay difference of a flight varies by comparing two time zones.

tbl <- xtabs(~ Time_Zone + Status, data = flight_data_long)
summary(tbl)
## Call: xtabs(formula = ~Time_Zone + Status, data = flight_data_long)
## Number of cases in table: 20 
## Number of factors: 2 
## Test for independence of all factors:
##  Chisq = 0, df = 1, p-value = 1
proportions(tbl, "Status")
##          Status
## Time_Zone delayed on_time
##   ALASKA      0.5     0.5
##   AM WEST     0.5     0.5

Conclusion

There is no difference between the two time zones regarding flight delays and punctuality. This was confirmed with a chi-squared test p-value = 1.