Mapping Accessibility to Health Food in Brazilian Favelas and Urban Comunities

Author

Marcio Antonio Pizzonia Junior

Published

June 2, 2026

Project Status: WIP – Initial development is in progress, but there has not yet been a stable, usable release suitable for the public. License: GPLv3 License: CC BY-NC-SA 4.0

Overview

The present study aims to provide maps generated through the Kernel Density Function regarding access to healthy food in Brazilian favelas and urban communities by obtaining, processing, geocoding, and classifying CNPJs data from the Brazilian Federal Revenue Service (RFB), as well as acquiring, processing, geocoding, and classifying (OSM) establishment data via the Overpass API.

The classification of food healthfulness follows Locais-Nova study, using Group 0 (G0), which is not part of the original classification but rather a subgroup within G1 + G2. This subgroup is composed of establishments that exclusively sell healthy foods that should constitute part of the population’s diet.

For instructions on how to run this work, see the repository README.

Problem

The AcessoSAN project aims to develop new methods for measuring and analyzing inequalities in access to healthy food, particularly in Brazilian favelas and urban communities. To achieve these objectives, it is necessary to use reliable and up-to-date datasets regarding establishments that sell healthy food.

The Brazilian Federal Revenue Service RFB provides records of companies operating in Brazil, including food retailers. However, it does not provide geocoded information, which is necessary for spatial analysis. This study follows the piepelnes developed by Daniel Vartanian et al. (2026b) to georeference RFB CNPJs data and by Flávio Soares de Freitas et al. (2026) obtain accessibility metrics for Group G0.

Furthermore, in order to better understand the existing informational gaps regarding access to healthy food in favelas and urban communities, analyses will also be conducted using data available through the Overpass API as a means of comparison with the results obtained from the RFB. RFB.

Data Availability

Methods

Source of Data

The data used for this this work comes from:

Data Munging

The data munging follows the data science workflow outlined by Wickham et al. (2023), as illustrated in Figure 1. All processes were made using the Quarto publishing system, along with the AWK (Aho et al., 2023) and R (R Core Team, n.d.) programming languages, supported by several R packages.

For data manipulation and workflow, priority was given to packages from the tidyverse, rOpenSci and r-spatial frameworks, as well as other packages adhering to the tidy tools manifesto (Wickham, 2023).

Figure 1: Data science workflow created by Wickham, Çetinkaya-Runde, and Grolemund.

Source: Reproduced from Wickham et al. (2023).

Data Structure

Brazilian Federal Revenue Service (RFB)

Open Street Map (OSM)

Travel Matrix

The travel matrix is computed using the r5r R package (Pereira et al., 2021), which interfaces with the R5 routing engine. This tool enables detailed travel time calculations based on multimodal transportation networks, accounting for road types, traffic conditions, and elevation changes. Travel times are derived from OpenStreetMap (OSM) road and footpath data (OpenStreetMap Contributors, n.d.) and elevation data from the Mapzen Terrain Tile dataset (Mapzen, n.d.).

Code Style

The Tidyverse tidy tools manifesto (Wickham, 2023), code style guide (Wickham, n.d.-a) and design principles (Wickham, n.d.-b) were followed to ensure consistency and enhance readability.

Reproducibility

The pipeline is fully reproducible and can be run again at any time. To ensure consistent results, the renv package (Ushey & Wickham, 2025) was used to manage and restore the R environment. See the README file in the code repository to learn how to run it.

Load Packages

Import, Process, Geocoded and Classify CNPJs from Bazilian Federal Revenue (RFB)

This pipeline was adapted from the pipeline made by Daniel Vartanian et al. (2026b) to geocode only the G0 groups.

Set the Environment

Set Keys

Veja a seção Disponibilidade dos Dados para mais informações.

osf_auth(Sys.getenv("OSF_PAT")) # askpass())
gs4_auth(
  cache = here(".secrets"),
  email = TRUE,
  use_oob = FALSE
)
public_key <- here(".ssh", "id_rsa.pub")
private_key <- here(".ssh", "id_rsa")
password <- Sys.getenv("ACESSOSAN_PASSWORD") # askpass()

Set Data Directories

raw_data_dir <- here("data-raw")
input_data_dir <- here("data-input")
data_dir <- here("data")
for (i in c(raw_data_dir, input_data_dir, data_dir)) {
  if (!dir_exists(i)) {
    dir_create(i, recurse = TRUE)
  }
}

Check Parameters

The pipeline main parameters are defined in the YAML header of its Quarto document.

The year and month variables represent the year and month of the consolidated RFB CNPJ dataset.

The municipality_codes variable identifies the IBGE codes of the municipalities included in the analysis. Click here to see a list of all Brazilian municipalities and their codes.

params |> lobstr::tree()
#> <list>
#> ├─year: 2026
#> ├─month: 1
#> ├─municipality_codes<int [5]>: 2507507, 3106200, 3550308, 4314902, 1501402
#> └─h3_resolution: 9
assert_int(params$year, lower = 1980, upper = 2050)
assert_int(params$month, lower = 1, upper = 12)
assert_integerish(params$municipality_codes, lower = 1100015, upper = 5300108)

Download and Import Locais-Nova CNAEs Lookup Table

See the Source of Data section for more information.

Although restaurants are part of the Locais-Nova classification, they were excluded from this analysis due to conflicting evidence on how to reliably identify them using company registry data.

Download Data

locais_nova_lookup_file <-
  input_data_dir |>
  file.path("locais-nova-lookup-data.rds")
if (!file_exists(locais_nova_lookup_file)) {
  locais_nova_lookup_data <-
    "1ipCw2FM3aUOdRd4w55J5SUEgXDCogxWZF-NZi0A-o_4" |>
    read_sheet(sheet = "Dataset")

  locais_nova_lookup_data |> write_rds(locais_nova_lookup_file)
}

Import Data

locais_nova_lookup_data <- locais_nova_lookup_file |> read_rds()
locais_nova_lookup_data |> glimpse()
#> Rows: 540
#> Columns: 7
#> $ cnae         <chr> "4711302", "5611201", "4722902", "4721102", "4721101",…
#> $ description  <chr> "Supermarket 1", "Restaurant", "Fish Market", "Bakery …
#> $ federal_unit <chr> "AC", "AC", "AC", "AC", "AC", "AC", "AC", "AC", "AC", …
#> $ g0           <lgl> FALSE, FALSE, TRUE, FALSE, FALSE, FALSE, FALSE, FALSE,…
#> $ g1_g2        <lgl> TRUE, TRUE, TRUE, FALSE, FALSE, FALSE, FALSE, FALSE, F…
#> $ g3           <lgl> FALSE, FALSE, FALSE, TRUE, TRUE, FALSE, FALSE, TRUE, F…
#> $ g4           <lgl> TRUE, FALSE, FALSE, TRUE, TRUE, TRUE, TRUE, TRUE, TRUE…

Download, Import, and Process Municipality Data

See the Source of Data section for more information.

Download and Save Data

read_municipality_file <-
  input_data_dir |>
  file.path(
    paste0(
      paste(
        "read_municipality",
        "code",
        "all",
        "year",
        params$year |> closest_geobr_year(type = "municipality"),
        "simplified",
        FALSE,
        sep = "-"
      ),
      ".rds"
    )
  )
#> ! The closest map year to 2026 is 2024. Using year 2024 instead.
if (!file_exists(read_municipality_file)) {
  municipalities_data <-
    read_municipality(
      year = params$year |> closest_geobr_year(type = "municipality"),
      simplified = FALSE,
      showProgress = FALSE
    )

  municipalities_data |> write_rds(read_municipality_file)
}

Import Data

municipalities_data <- read_municipality_file |> read_rds()
municipalities_data |> glimpse()
#> Rows: 5,571
#> Columns: 8
#> $ code_muni    <dbl> 1100015, 1100023, 1100031, 1100049, 1100056, 1100064, …
#> $ name_muni    <chr> "Alta Floresta D'oeste", "Ariquemes", "Cabixi", "Cacoa…
#> $ code_state   <dbl> 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11…
#> $ abbrev_state <chr> "RO", "RO", "RO", "RO", "RO", "RO", "RO", "RO", "RO", …
#> $ name_state   <chr> "Rondônia", "Rondônia", "Rondônia", "Rondônia", "Rondô…
#> $ code_region  <dbl> 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, …
#> $ name_region  <chr> "Norte", "Norte", "Norte", "Norte", "Norte", "Norte", …
#> $ geom         <MULTIPOLYGON [°]> MULTIPOLYGON (((-61.9300418..., MULTIPOLY…

Filter Data

municipalities_data <-
  municipalities_data |>
  filter(code_muni %in% params$municipality_codes)
municipalities_data |> glimpse()
#> Rows: 5
#> Columns: 8
#> $ code_muni    <dbl> 1501402, 2507507, 3106200, 3550308, 4314902
#> $ name_muni    <chr> "Belém", "João Pessoa", "Belo Horizonte", "São Paulo",…
#> $ code_state   <dbl> 15, 25, 31, 35, 43
#> $ abbrev_state <chr> "PA", "PB", "MG", "SP", "RS"
#> $ name_state   <chr> "Pará", "Paraíba", "Minas Gerais", "São Paulo", "Rio G…
#> $ code_region  <dbl> 1, 2, 3, 3, 4
#> $ name_region  <chr> "Norte", "Nordeste", "Sudeste", "Sudeste", "Sul"
#> $ geom         <MULTIPOLYGON [°]> MULTIPOLYGON (((-48.5257946..., MULTIPOLYGON (((-34.88…

Transform Data

municipalities_data <-
  municipalities_data |>
  st_transform(st_crs(4326)) |>
  select(
    municipality = name_muni,
    municipality_code = code_muni,
    state = name_state,
    state_code = code_state,
    federal_unit = abbrev_state,
    region = name_region,
    region_code = code_region,
    geometry = geom
  )
municipalities_data |> glimpse()
#> Rows: 5
#> Columns: 8
#> $ municipality      <chr> "Belém", "João Pessoa", "Belo Horizonte", "São Pa…
#> $ municipality_code <dbl> 1501402, 2507507, 3106200, 3550308, 4314902
#> $ state             <chr> "Pará", "Paraíba", "Minas Gerais", "São Paulo", "…
#> $ state_code        <dbl> 15, 25, 31, 35, 43
#> $ federal_unit      <chr> "PA", "PB", "MG", "SP", "RS"
#> $ region            <chr> "Norte", "Nordeste", "Sudeste", "Sudeste", "Sul"
#> $ region_code       <dbl> 1, 2, 3, 3, 4
#> $ geometry          <MULTIPOLYGON [°]> MULTIPOLYGON (((-48.5257946..., MULTIPOLYGON (((-…

Download, Import, and Transform Agencies and Municipalities (TOM) Data

See the Source of Data section for more information.

Download Data

municipalities_tom_data_file <-
  input_data_dir |>
  file.path("municipalities_tom_data.csv")
if (!file_exists(municipalities_tom_data_file)) {
  request("https://www.gov.br") |>
    req_url_path_append("receitafederal") |>
    req_url_path_append("dados") |>
    req_url_path_append("municipios.csv") |>
    req_user_agent("https://cem-usp.github.io/kernel-accessibility") |>
    req_options(connecttimeout = 10) |>
    req_retry(
      max_tries = 3,
      retry_on_failure = TRUE,
      backoff = \(attempt) 5^attempt
    ) |>
    req_progress() |>
    req_perform(path = municipalities_tom_data_file)
}

Import Data

municipalities_tom_data <-
  municipalities_tom_data_file |>
  read_delim(
    delim = ";",
    col_names = FALSE,
    col_types = cols(.default = "c"),
    progress = FALSE
  ) |>
  slice(-1)
municipalities_tom_data |> glimpse()
#> Rows: 5,571
#> Columns: 5
#> $ X1 <chr> "1", "2", "3", "4", "5", "6", "7", "8", "9", "10", "11", "12", "…
#> $ X2 <chr> "1100106", "1100379", "1100205", "1100452", "1100122", "1100924"…
#> $ X3 <chr> "GUAJAR\xc1-MIRIM", "ALTO ALEGRE DOS PARECIS", "PORTO VELHO", "B…
#> $ X4 <chr> "Guajar\xe1-Mirim", "Alto Alegre dos Parecis", "Porto Velho", "B…
#> $ X5 <chr> "RO", "RO", "RO", "RO", "RO", "RO", "RO", "RO", "RO", "RO", "RO"…

Transform Data

municipalities_tom_data <-
  municipalities_tom_data |>
  mutate(
    across(
      .cols = everything(),
      .fns = \(x) iconv(x, from = "Windows-1252", to = "UTF-8") # iconvlist()
    )
  ) |>
  set_names(
    c(
      "municipality_code_tom",
      "municipality_code_ibge",
      "municipality_tom",
      "municipality_ibge",
      "federal_unit"
    )
  ) |>
  mutate(
    across(
      .cols = starts_with("municipality_code"),
      .fns = as.integer
    )
  ) |>
  relocate(
    municipality_tom,
    municipality_code_tom,
    municipality_ibge,
    municipality_code_ibge,
    federal_unit
  )
municipalities_tom_data |> glimpse()
#> Rows: 5,571
#> Columns: 5
#> $ municipality_tom       <chr> "GUAJARÁ-MIRIM", "ALTO ALEGRE DOS PARECIS", …
#> $ municipality_code_tom  <int> 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 1…
#> $ municipality_ibge      <chr> "Guajará-Mirim", "Alto Alegre dos Parecis", …
#> $ municipality_code_ibge <int> 1100106, 1100379, 1100205, 1100452, 1100122,…
#> $ federal_unit           <chr> "RO", "RO", "RO", "RO", "RO", "RO", "RO", "R…

Map IBGE Municipality Codes to TOM Codes

The Brazilian Federal Revenue Service (RFB) identifies municipalities using a different coding scheme than IBGE codes (Tabela de Órgãos e Municípios, TOM). Mapping IBGE codes to the corresponding TOM codes is therefore required.

params$municipality_codes
#> [1] 2507507 3106200 3550308 4314902 1501402
municipality_tom_codes <-
  municipalities_tom_data |>
  filter(municipality_code_ibge %in% params$municipality_codes) |>
  pull(municipality_code_tom) |>
  str_pad(width = 4, pad = "0")
municipality_tom_codes
#> [1] "0427" "2051" "4123" "7107" "8801"

Download RFB Data

See the Source of Data section for more information.

List Files

public_token <- "YggdBLfdninEJX9"
request <-
  request("https://arquivos.receitafederal.gov.br") |>
  req_auth_basic(public_token, "")
response <-
  request |>
  req_url_path_append("public.php") |>
  req_url_path_append("webdav") |>
  req_method("PROPFIND") |>
  req_url_path_append(
    paste0(params$year, "-", str_pad(params$month, 2, pad = 0))
  ) |>
  req_headers("Depth" = "1") |>
  req_body_raw(
    '<?xml version="1.0"?>
      <d:propfind xmlns:d="DAV:">
        <d:prop>
          <d:displayname/>
          <d:getcontentlength/>
          <d:getcontenttype/>
          <d:getlastmodified/>
        </d:prop>
      </d:propfind>',
    type = "application/xml"
  ) |>
  req_user_agent("https://cem-usp.github.io/kernel-accessibility") |>
  req_options(connecttimeout = 10) |>
  req_retry(
    max_tries = 3,
    retry_on_failure = TRUE,
    backoff = \(attempt) 5^attempt
  ) |>
  req_perform() |>
  resp_body_xml() %>%
  xml_find_all(
    .,
    ".//d:response",
    xml_ns(.)
  )
list_files <- function(node, response) {
  xpaths <- list(
    name = ".//d:displayname",
    url = ".//d:href",
    size = ".//d:getcontentlength",
    type = ".//d:getcontenttype",
    modified = ".//d:getlastmodified"
  )

  xpaths |>
    map(
      function(xpath) {
        node |>
          xml_find_first(xpath, xml_ns(response)) |>
          xml_text()
      }
    ) |>
    as_tibble()
}
files <-
  response |>
  map(list_files, response = response) |>
  list_rbind() |>
  filter(type == "application/zip") |>
  mutate(size = size |> as_fs_bytes())
files |> glimpse()
#> Rows: 37
#> Columns: 5
#> $ name     <chr> "Cnaes.zip", "Empresas0.zip", "Empresas1.zip", "Empresas2.…
#> $ url      <chr> "/public.php/webdav/2026-01/Cnaes.zip", "/public.php/webda…
#> $ size     <fs::bytes> 21.56K, 465.72M, 74.25M, 75.43M, 81.18M, 86.12M, 92.…
#> $ type     <chr> "application/zip", "application/zip", "application/zip", "…
#> $ modified <chr> "Sun, 11 Jan 2026 17:51:31 GMT", "Sun, 11 Jan 2026 17:52:0…

Download Data

rfb_file_pattern <-
  paste0(
    "D",
    params$year |> str_sub(-1),
    params$month |> str_pad(width = 2, pad = "0")
  )
if (
  raw_data_dir |>
    dir_ls() |>
    str_detect(rfb_file_pattern) |>
    all() |>
    not()
) {
  cli_progress_bar(
    "Downloading RFB files",
    total = nrow(files),
    clear = FALSE
  )

  for (i in seq_len(nrow(files))) {
    request |>
      req_url_path_append(files$url[i]) |>
      req_perform(
        path = here(raw_data_dir, files$name[i])
      )

    cli_progress_update()
  }

  cli_progress_done()
}

Unzip Files

zip_files <-
  raw_data_dir |>
  dir_ls(type = "file", regexp = "\\.zip$")
if (length(zip_files) > 0) {
  zip_files |>
    map(
      \(x) unzip(x, exdir = raw_data_dir),
      .progress = TRUE
    )
}
if (length(zip_files) > 0) {
  zip_files |> file_delete()
}

Import and Filter Data

Import and Filter Active and Regular Establishments by Municipalities

The AWK programming language and the vroom R package are used to read and filter the data efficiently, without loading it entirely into memory.

Only establishments with an Active registration status ($6 == "02") and no special status ($29 == "") are included, and the resulting records are further filtered by municipality ($21 == municipality_tom_codes).

Click here to access the data dictionary.

data <-
  raw_data_dir |>
  dir_ls(
    type = "file",
    regexp = "\\.ESTABELE$"
  ) |>
  map(
    function(x) {
      vroom(
        file = pipe(
          paste0(
            "awk ",
            "-F ",
            "';' ",
            "'{ ",
            "if (",
            '($6 == "',
            '\\"02\\"',
            '")',
            " && (",
            paste0(
              '$21 == "\\"',
              municipality_tom_codes,
              '\\""',
              collapse = " || "
            ),
            ") && ",
            '($29 == "',
            '\\"\\"',
            '")',
            ") ",
            "print",
            " }' ",
            str_escape(x)
          )
        ),
        delim = ";",
        col_names = c(
          "cnpj_basic",
          "cnpj_order",
          "cnpj_dv",
          "branch_identifier",
          "trade_name",
          "registration_status",
          "registration_status_date",
          "registration_status_reason",
          "foreign_city_name",
          "country",
          "opening_date",
          "primary_cnae",
          "secondary_cnae",
          "street_type",
          "street_name",
          "number",
          "complement",
          "neighborhood",
          "postal_code",
          "federal_unit",
          "municipality_code_tom",
          "phone_area_code_1",
          "phone_number_1",
          "phone_area_code_2",
          "phone_number_2",
          "fax_area_code",
          "fax_number",
          "email",
          "special_status",
          "special_status_date"
        ),
        col_types = cols(.default = "c"),
        col_select = c(
          "cnpj_basic",
          "cnpj_order",
          "cnpj_dv",
          "trade_name",
          "opening_date",
          "primary_cnae",
          "secondary_cnae",
          "street_type",
          "street_name",
          "number",
          "complement",
          "neighborhood",
          "postal_code",
          "municipality_code_tom",
          "federal_unit",
          "registration_status_date"
        ),
        guess_max = 100,
        num_threads = detectCores() |>
          multiply_by(0.75) |>
          ceiling(),
        progress = TRUE,
        show_col_types = FALSE
      )
    },
    .progress = TRUE
  ) |>
  list_rbind()
data |> glimpse()
#> Rows: 0
#> Columns: 16
#> $ cnpj_basic               <chr> 
#> $ cnpj_order               <chr> 
#> $ cnpj_dv                  <chr> 
#> $ trade_name               <chr> 
#> $ opening_date             <chr> 
#> $ primary_cnae             <chr> 
#> $ secondary_cnae           <chr> 
#> $ street_type              <chr> 
#> $ street_name              <chr> 
#> $ number                   <chr> 
#> $ complement               <chr> 
#> $ neighborhood             <chr> 
#> $ postal_code              <chr> 
#> $ municipality_code_tom    <chr> 
#> $ federal_unit             <chr> 
#> $ registration_status_date <chr>

Filter Establishments by Locais-Nova G0 Group

data <-
  data |>
  mutate(primary_cnae = primary_cnae |> str_pad(7, pad = "0")) |>
  inner_join(
    y = locais_nova_lookup_data,
    by = join_by(primary_cnae == cnae, federal_unit)
  ) |>
  rename(
    locais_nova_g0 = g0,
    locais_nova_g1_g2 = g1_g2,
    locais_nova_g3 = g3,
    locais_nova_g4 = g4
  ) 
data |> glimpse()
#> Rows: 0
#> Columns: 21
#> $ cnpj_basic               <chr> 
#> $ cnpj_order               <chr> 
#> $ cnpj_dv                  <chr> 
#> $ trade_name               <chr> 
#> $ opening_date             <chr> 
#> $ primary_cnae             <chr> 
#> $ secondary_cnae           <chr> 
#> $ street_type              <chr> 
#> $ street_name              <chr> 
#> $ number                   <chr> 
#> $ complement               <chr> 
#> $ neighborhood             <chr> 
#> $ postal_code              <chr> 
#> $ municipality_code_tom    <chr> 
#> $ federal_unit             <chr> 
#> $ registration_status_date <chr> 
#> $ description              <chr> 
#> $ locais_nova_g0           <lgl> 
#> $ locais_nova_g1_g2        <lgl> 
#> $ locais_nova_g3           <lgl> 
#> $ locais_nova_g4           <lgl>

Import and Filter Additional Company Data by Establishment

Import and Filter Simples Nacional Data by Establishment

Visualize Registration Status Dates

These dates indicate when establishments were registered as Active.

Tidy Data

Some tidying operations are performed to ensure the data is in a suitable format for analysis. The data is not yet reshaped to a tidy format to avoid duplicated cases during geocoding.

Addresses were normalized using the enderecobr R package (Herszenhut et al., n.d.), which standardizes common street types and abbreviations in Brazilian addresses.

Remove Extra Columns

Standardize Character Encoding

Pad CNPJs and CNAEs

Remove CNPJ Duplicates

Remove Invalid CNPJs and CNAEs

Add Municipality Data

Standardize Data

Relocate Columns

Arrange Data

Geocode Data

Prepare Data

Geocode Data

Tidy Data

Visualize Geocoding Precision

Global Precision
Local Precision

Tidy Data

Remove Extra Columns

Pivot Data

Arrange Data

Create Data Dictionary

Prepare Metadata

Visualize Final Data

Save Data

The processed data are available in csv, rds and parquet formats through a dedicated repository on the Open Science Framework (OSF). See the Data Availability section for more information.

Clean Old Data Files

Write Data

Write Lookup Tables

Lock Data

Write Meetadata

Visualize Geocoding

The processed data can be visualized at RFB and OSM Georeferenced Data.

Open Street Map (OSM) Establishment Data Acquisition, Processing, Geocoding, and Classification

This pipeline was developed to establish a methodology for assessing informational gaps in data availability within Brazilian favelas and urban communities.

RFB and OSM Georeferenced Data

Visualize the RFB and OSM georeferenced data.

Calculating Accessibility to Food Establishments in Brazilian Favelas and Urban Comunities

This pipeline was adapted from the pipeline made by Flávio Soares de Freitas et al. (2026) to calculate the accessibility to food establishments classified as group G0 in favelas and urban comunities.

Furthermore, this pipeline uses the RFB and OSM data to obtain the metrics.

Mapping Acessibility to Food Establishments in Brazilian Favelas and Urban Comunities

This pipeline was adapted from the pipeline made by Daniel Vartanian et al. (2026a) to map the accessibility to food establishments classified as group G0 in favelas and urban comunities.

Furthermore, this pipeline uses the accessibility metrics obtained from the Calculating Accessibility to Food Establishments in Brazilian Favelas and Urban Comunities.

Mapping the Accessibility Denisty in Brazilian Favelas and Urban Comunities

This pipeline offers the accesibility density trough the Kernel Density Function.

As the Mapping Acessibility to Food Establishments in Brazilian Favelas and Urban Comunities, this pipeline also uses the metrics obtained from the Calculating Accessibility to Food Establishments in Brazilian Favelas and Urban Comunities.

Outcomes

Analisys of the Outcomes

Conclusion

Citation

When using this data, you must also cite the original data sources.

To cite this work, please use the following format:

Pizzonia, M. A. Jr. (2026). Mapping Accessibility to Health Food in Brazilian Favelas and Urban Comunities [Computer software]. Center for Metropolitan Studies, University of São Paulo. https://cem-usp.github.io/kernel-accessibility/

A BibTeX entry for LaTeX users is

@software{pizzonia2026,
  title = {Mapping Healthy Food Accessibility in Brazilian Slums and Urban Communities},
  author = {{Marcio Antonio Pizzonia Junior}},
  year = {2026},
  address = {São Paulo},
  institution = {Center for Metropolitan Studies, University of São Paulo},
  langid = {en},
  url = {https://cem-usp.github.io/kernel-accessibility/}
}

License

License: GPLv3 License: CC BY-NC-SA 4.0

The original data sources may be subject to their own licensing terms and conditions.

The code in this repository is licensed under the GNU General Public License Version 3, while the report is available under the Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International license.

Copyright (C) 2026 Center for Metropolitan Studies

The code in this report is free software: you can redistribute it and/or
modify it under the terms of the GNU General Public License as published by the
Free Software Foundation, either version 3 of the License, or (at your option)
any later version.

This program is distributed in the hope that it will be useful, but WITHOUT ANY
WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A
PARTICULAR PURPOSE. See the GNU General Public License for more details.

You should have received a copy of the GNU General Public License along with
this program. If not, see <https://www.gnu.org/licenses/>.

Reconhecimentos

This work is part of a research project by the Polytechnic School (Poli) of the University of São Paulo (USP), in partnership with the National Secretariat for Food and Nutrition Security (SESAN) of the Brazilian Ministry of Social Development and Assistance, Family and Fight Against Hunger (MDS), titled: AcessoSAN: Mapping Food Access to Support Public Policies on Food and Nutrition Security and Hunger Reduction in Brazilian Cities.

This work was developed with support from the Center for Metropolitan Studies (CEM) based at the School of Philosophy, Letters and Human Sciences (FFLCH) of the University of São Paulo (USP) and at the Brazilian Center for Analysis and Planning (CEBRAP).

This study was financed, in part, by the São Paulo Research Foundation (FAPESP), Brazil. Process Number 2025/17879-2.

Referencias

Aho, A., Kernighan, B., & Weinberger, P. (2023). The AWK programming language. Addison-Wesley Professional. https://www.awk.dev
Brasil, Ministério da Saúde, & Departamento de Informática do SUS. (n.d.). Banco de dados do Sistema Único de Saúde - DATASUS. Retrieved July 17, 2024, from https://datasus.saude.gov.br/
Brasil, & Sistema de Vigilância Alimentar e Nutricional. (n.d.). Módulo gerador de relatórios públicos: estado nutricional dos indivíduos acompanhados por período, fase do ciclo da vida e índice. Retrieved September 23, 2024, from https://sisaps.saude.gov.br/sisvan/relatoriopublico
Daniel Vartanian, Gabriel Pereira Caldeira, Clara de Lima e Silva Penz, Camila Nastari Fernandes, & Mariana Giannotti. (2026a). Reproducible pipeline for mapping accessibility to food establishments in brazilian municipalities using the locais-nova classification [Computer software]. Center for Metropolitan Studies, University of São Paulo. https://cem-usp.github.io/locais-nova-mapping
Daniel Vartanian, Gabriel Pereira Caldeira, Clara de Lima e Silva Penz, Camila Nastari Fernandes, & Mariana Giannotti. (2026b). Reproducible pipeline for processing, geocoding, and classifying CNPJs from the brazilian federal revenue service (RFB) using the locais-nova classification [Computer software]. Center for Metropolitan Studies, University of São Paulo. https://cem-usp.github.io/locais-nova-rfb-geocoding
Fenandes, C. N., & Giannotti, M. (2026). Lookup table linking establishment CNAE codes to their corresponding Locais-Nova groups [Data set].
Flávio Soares de Freitas, Clara de Lima e Silva Penz, Daniel Vartanian, Camila Nastari Fernandes, & Mariana Abrantes Giannotti. (2026). Reproducible pipeline for calculating accessibility to food establishments in brazilian municipalities using the locais-nova classification [Computer software]. Center for Metropolitan Studies, University of São Paulo. https://cem-usp.github.io/locais-nova-access
Geofabrik GmbH. (n.d.). OpenStreetMap for Brazil [Data set]. https://download.geofabrik.de/south-america/brazil.html
Herszenhut, D., Pereira, R. H. M., & Mation, L. (n.d.). enderecobr: Brazilian addresses standardizer [Computer software]. https://doi.org/10.32614/CRAN.package.enderecobr
Hollister, J., Shah, T., Nowosad, J., Robitaille, A. L., Beck, M. W., & Johnson, M. (2025). elevatr: Access elevation data from various APIs [Computer software]. https://doi.org/10.32614/CRAN.package.elevatr
Instituto Brasileiro de Geografia e Estatística, & Comissão Nacional de Classificação. (n.d.). CNAE 2.3 Subclasses [Data set]. SIDRA. Retrieved November 16, 2023, from https://concla.ibge.gov.br/classificacoes/download-concla.html
MapLibre Contributors. (2022). MapLibre GL JS [Computer software]. https://maplibre.org
Mapzen. (n.d.). Terrain tiles [Data set]. https://registry.opendata.aws/terrain-tiles
O’Brien, L. (2023). h3jsr: Access Uber’s H3 library [Computer software]. https://doi.org/10.32614/CRAN.package.h3jsr
OpenStreetMap Contributors. (n.d.). OpenStreetMap [Computer software]. OpenStreetMap Foundation. https://www.openstreetmap.org
Pereira, R. H. M., & Goncalves, C. N. (n.d.). geobr: Download official spatial data sets of Brazil [Computer software]. https://doi.org/10.32614/CRAN.package.geobr
Pereira, R. H. M., & Herszenhut, D. (n.d.). geocodebr: Geolocalização de endereços brasileiros [geocodebr: Geolocation of Brazilian addresses] [Computer software]. https://doi.org/10.32614/CRAN.package.geocodebr
Pereira, R. H. M., Saraiva, M., Herszenhut, D., Braga, C. K. V., & Conway, M. W. (2021). r5r: Rapid Realistic Routing on Multimodal Transport Networks with R5 in R. Findings. https://doi.org/10.32866/001c.21262
R Core Team. (n.d.). R: A language and environment for statistical computing [Computer software]. R Foundation for Statistical Computing. https://www.r-project.org
Uber Technologies. (n.d.). H3: A hexagonal hierarchical geospatial indexing system [Data set]. https://h3geo.org
Ushey, K., & Wickham, H. (2025). renv: Project environments [Computer software]. https://doi.org/10.32614/CRAN.package.renv
Vartanian, D., Fernandes, C. N., & Giannotti, M. (2026). PMTilesBR: Tiled geospatial data for Brazil [Computer software]. Center for Metropolitan Studies, University of São Paulo. https://doi.org/10.5281/zenodo.19157888
Walker, K. (2026). mapgl: Interactive Maps with ’Mapbox GL JS’ and ’MapLibre GL JS’ [Computer software]. https://walker-data.com/mapgl
Wickham, H. (n.d.-a). The tidyverse style guide. https://style.tidyverse.org
Wickham, H. (n.d.-b). Tidy design principles. https://design.tidyverse.org
Wickham, H. (2023). The tidy tools manifesto. Tidyverse. https://tidyverse.tidyverse.org/articles/manifesto.html
Wickham, H., Çetinkaya-Rundel, M., & Grolemund, G. (2023). R for data science: Import, tidy, transform, visualize, and model data (2nd ed.). O’Reilly Media. https://r4ds.hadley.nz