Skip to contents

Getting Started

Source: vignettes/ostutils.Rmd
ostutils.Rmd

The ost.utils package provides tools and methods to streamline the workflow in R projects developed by the Coordenadoria do Observatório de Segurança no Trânsito (COST) of Detran-SP.

Installation

The development version of ost.utils can be installed from GitHub with:

# install.packages("pak")
pak::pak("pedrobsantos21/ost.utils")

Package organization

This package is organized into two main groups of functions:

  1. infosiga: methods to download, load and clean open data from Infosiga.SP
  1. plot: helper functions to plot data with ggplot2:

Usage example

In this example, we will load Infosiga road crash data and plot it using ggplot2. First, we load the required packages:

library(ost.utils)
library(ggplot2)
library(dplyr)
#> 
#> Attaching package: 'dplyr'
#> The following objects are masked from 'package:stats':
#> 
#>     filter, lag
#> The following objects are masked from 'package:base':
#> 
#>     intersect, setdiff, setequal, union
library(lubridate)
#> 
#> Attaching package: 'lubridate'
#> The following objects are masked from 'package:base':
#> 
#>     date, intersect, setdiff, union

Then, we use download_infosiga() to save the data to a temporary folder, load the road crash data with load_infosiga(), and clean it with clean_infosiga(). In a typical project, you might download the data to a dedicated data/ folder.

temp <- tempdir()
download_infosiga(temp)
#> ℹ Starting download...
#> ✔ Download completed.
#> ℹ Extrating zip...
#> ✔ Data extracted successfully at '/tmp/RtmpmHdmfT'

df <- load_infosiga(file_type = "sinistros", path = temp)
#> ℹ Using "','" as decimal and "'.'" as grouping mark. Use `read_delim()` for more control.
#> Rows: 1208097 Columns: 43
#> ── Column specification ────────────────────────────────────────────────────────
#> Delimiter: ";"
#> chr  (26): tipo_registro, data_sinistro, mes_sinistro, dia_sinistro, ano_mes...
#> dbl  (15): id_sinistro, ano_sinistro, latitude, longitude, tp_veiculo_bicicl...
#> lgl   (1): gravidade_ileso
#> time  (1): hora_sinistro
#> 
#> ℹ Use `spec()` to retrieve the full column specification for this data.
#> ℹ Specify the column types or set `show_col_types = FALSE` to quiet this message.

df_clean <- clean_infosiga(df, file_type = "sinistros")

head(df_clean)
#> # A tibble: 6 × 40
#>   id_sinistro data_sinistro hora_sinistro cod_ibge regiao_administrativa     
#>         <dbl> <date>        <time>        <chr>    <chr>                     
#> 1     2501575 2014-12-21    20:00         3509502  Campinas                  
#> 2     2456933 2014-12-23       NA         3505500  Barretos                  
#> 3     2463759 2014-12-26    06:52         3550308  Metropolitana de São Paulo
#> 4     2487781 2014-12-28    14:30         3510609  Metropolitana de São Paulo
#> 5     2489730 2014-12-28       NA         3541000  Baixada Santista          
#> 6     2462674 2014-12-31    22:53         3550308  Metropolitana de São Paulo
#> # ℹ 35 more variables: nome_municipio <chr>, logradouro <chr>,
#> #   numero_logradouro <dbl>, tipo_via <chr>, longitude <dbl>, latitude <dbl>,
#> #   tp_veiculo_bicicleta <dbl>, tp_veiculo_caminhao <dbl>,
#> #   tp_veiculo_motocicleta <dbl>, tp_veiculo_nao_disponivel <dbl>,
#> #   tp_veiculo_onibus <dbl>, tp_veiculo_outros <dbl>,
#> #   tp_veiculo_automovel <dbl>, tipo_registro <chr>,
#> #   gravidade_nao_disponivel <dbl>, gravidade_leve <dbl>, …

Now we can plot the count of road crashes per year using the custom Detran style:

df_clean |> 
  filter(
    tipo_registro %in% c("Sinistro fatal", "Sinistro não fatal"),
    year(data_sinistro) > 2018
  ) |> 
  count(year = year(data_sinistro)) |> 
  ggplot(aes(x = year, y=n)) +
  geom_col(fill = palette_detran()$blue) +
  theme_detran()