Comparing air passenger demand over time in Brazil

air_passenger_demand_BR.R

#' Comparing air passenger demand over time in Brazil
#' Reproducible code to create a figure visualizing how daily air passenger demand in Brazil has changed over time.


### Load libraries
library(ggplot2)
library(flightsbr)
library(lubridate)
library(data.table)

### Download and rbind data
d2019 <- read_flights(date = 2019)
d2020 <- read_flights(date = 2020)
d2021 <- read_flights(date = 2021)
d2022 <- read_flights(date = 2022)

df <- rbindlist(list(d2019, d2020, d2021, d2022))


### Recode columns and sum daily number of passengers
df[, date := lubridate::as_date(dt_chegada_real)]
df[, year := lubridate::year(date)]
df[, month := lubridate::month(date)]
df[, day := lubridate::day(date)]
df[, day_week := lubridate::wday(date, label = TRUE)]

# sum total of passenger by day and flight type
df2 <- df[, .(total_pass = sum(nr_passag_pagos, nr_passag_gratis, na.rm=T)), by=.(date, year)]
df2[, xx := paste0("2020-", format(date, "%m-%d"))]
df2[, xx := lubridate::as_date(xx, format="%Y-%m-%d") ]
df2[, month := lubridate::month(date) ]
head(df2)

df2 <- subset(df2, year %in% c(2019, 2020, 2021, 2022))
df2 <- subset(df2, date < as.Date('2022-01-31'))


### Figure
fig <- ggplot() + 
        # geom_smooth( data= df2, aes(x=xx, y=total_pass/1000, color=factor(year)), alpha=.5, size=1) +
        geom_point( data= df2, aes(x=xx, y=total_pass/1000, color=factor(year)), alpha=.5, size=1) +
        scale_x_date(date_breaks = "1 months", date_labels = "%b") +
        labs(y='N. of passengers\n(thousands)', color='Year', x= 'Month',
             title ="Daily air passenger demand, Brazil", caption = 'figure by @UrbanDemog') +
        scale_colour_manual(values = c('gray70', 'gray30', 'black', 'red')) +
        theme_classic() +
        theme(plot.caption=element_text(colour="gray"))

ggsave(fig, filename = 'passenger_demand.png', dpi = 300, 
       width = 18, height = 10, units = 'cm')