Flight

Import Data

flight <- read.csv("flight.csv")

Remove blank row

flightdf <- flight[-c(3),]

Replace missing values

flightdf$X[2] = 'Alaska'
flightdf$X[4] = 'AM WEST'

Rename columns

names(flightdf) <- c('airline', 'status', 'Los.Angeles', 'Phoenix', 'San.Diego', ' San.Francisco', 'Seattle')

Convert from wide to long data

long <- gather(flightdf, state, measurement, Los.Angeles:Seattle, factor_key = T)

Find average delay by state

delay_state <- long %>% filter(status == 'delayed') %>%
  group_by(airline, state) %>%
  summarise(avg_delay = mean(measurement, na.rm = TRUE)) %>%
  arrange(avg_delay)

Plot data

ggplot(delay_state, aes(x=airline, y=avg_delay, fill=state)) +
geom_col(position = "dodge") +
labs(x = "Airline", y = "Average delays")

Conclusion

The graph above reflects that on average AM West has a significant delay in flights at Phoenix and Alaska airline does not. Alaska airline has a significant delay in flights leaving from Seattle and AM west airline does not.