---
title: "Flights data"
output: rmarkdown::html_vignette
vignette: >
  %\VignetteIndexEntry{Flights data}
  %\VignetteEngine{knitr::rmarkdown}
  %\VignetteEncoding{UTF-8}
---

```{r, include = FALSE}
knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>"
)
```

The flight data in the **flightsbr** package is downloaded from Brazil’s Civil Aviation Agency (ANAC). The data includes detailed information on every international flight to and from Brazil, as well as domestic flights within the country. The data include flight-level information of airports of origin and destination, flight duration, aircraft  type, payload, and the number of passengers, and several other variables. 

- Data dictionary: [a description of all variables included in the data is available here](https://www.gov.br/anac/pt-br/assuntos/regulados/empresas-aereas/Instrucoes-para-a-elaboracao-e-apresentacao-das-demonstracoes-contabeis/descricao-de-variaveis).


Now we can load some libraries we'll use in this vignette:

```{r, eval=FALSE, message = FALSE}
library(flightsbr)
library(data.table)
library(ggplot2)

```

Download data of all flights:

```{r, eval=FALSE}
# in a given **month* of a given **year** (yyyymm)
df_201506 <- read_flights(date = 201506)


# from specific months
df_various_months <- read_flights(date = c(202001, 202101, 202210))


# in a given year (yyyy)
df_2015 <- read_flights(date = 2015)


# from specific years
df_various_years <- read_flights(date = c(2019, 2021, 2022))

```


If you know already what data columns you need, you can pass a vector with their names to the `select` argument and `read_flights()` will only load those columns. This will make the function a bit faster. 

```{r, eval=FALSE}
df_201506 <- read_flights(
  date = 201506,
  showProgress = FALSE,
  select = c('id_empresa', 'nr_voo', 'dt_partida_real',
             'sg_iata_origem' , 'sg_iata_destino')
  )

head(df_201506)
```

The package makes it easy to compare daily number of passengers across different years. In the example below we compare daily number of air passengers in Brazil in 2019 and 2020. This gives us a glimpse in the impact of COVID-19 on Brazilian aviation, similarly to study of [Bazzo, Braga and Pereira (2022)](https://doi.org/10.1016/j.enpol.2022.112906).


```{r, eval=FALSE}
# download flights data
df <- read_flights(
  date = 2019:2022,
  select = c('nr_passag_pagos', 'dt_partida_real'),
  showProgress = TRUE
  )

# count daily passengers
count_df <- df[, .(total_pass = sum(nr_passag_pagos, na.rm=TRUE)), 
               by = dt_partida_real]

# reformat date
count_df <- count_df[ between(dt_partida_real, as.Date('2019-01-01'), as.Date('2022-12-31')) ]

count_df[, date := as.IDate(dt_partida_real, format="%Y-%m-%d") ]
count_df[, year := year(date) ]
count_df[, date_plot := paste0("2030-", format(date, "%m-%d"))]
count_df[, date_plot := as.Date(date_plot)]

# plot
fig <- ggplot(data = count_df) + 
          geom_point(aes(x=date_plot, y=total_pass, color=factor(year)), alpha=.4, size=1) +
          scale_y_log10(name="Number of Passengers", 
                        labels = scales::unit_format(unit = ""), limit=c(1000,NA)) +
          scale_x_date(date_breaks = "1 months", date_labels = "%b", name = 'Month') +
          labs(subtitle ='Daily number of air passengers in Brazil', color = "Legend") +
          theme_minimal() +
          theme(panel.grid.minor = element_blank(),
                axis.text = element_text(size = 7),
                axis.title=element_text(size=9),
                plot.background = element_rect(fill='white', colour='white'))


fig

```

```{r daily passengers, eval=TRUE, echo=FALSE, message=FALSE, out.width='100%'}
knitr::include_graphics("https://github.com/ipeaGIT/flightsbr/blob/main/inst/img/vig_output_flights.png?raw=true")
```