Skip to contents

Bags: using boundaries from NSO

Source: vignettes/articles/bags-from-nso.Rmd
bags-from-nso.Rmd

Bags (baga) are the smallest rural units: each soum is divided into a few bags, about 1,650 in all. mongolmaps knows every bag’s code and name:

mn_codes("bag", within = "Khovd")
#> # A tibble: 91 × 16
#>    pcode    name   name_en name_mn name_mns level type  number iso_code nso_code
#>    <chr>    <chr>  <chr>   <chr>   <chr>    <chr> <chr>  <int> <chr>    <chr>   
#>  1 MN840151 1-r b… 1-r ba… 1-р ба… 1-r bag… bag   bag        1 NA       1840151 
#>  2 MN840153 2-r b… 2-r ba… 2-р ба… 2-r bag… bag   bag        2 NA       1840153 
#>  3 MN840155 3-r b… 3-r ba… 3-р ба… 3-r bag… bag   bag        3 NA       1840155 
#>  4 MN840157 4-r b… 4-r ba… 4-р ба… 4-r bag… bag   bag        4 NA       1840157 
#>  5 MN840159 5-r b… 5-r ba… 5-р ба… 5-r bag… bag   bag        5 NA       1840159 
#>  6 MN840161 6-r b… 6-r ba… 6-р ба… 6-r bag… bag   bag        6 NA       1840161 
#>  7 MN840163 7-r b… 7-r ba… 7-р ба… 7-r bag… bag   bag        7 NA       1840163 
#>  8 MN840165 8-r b… 8-r ba… 8-р ба… 8-r bag… bag   bag        8 NA       1840165 
#>  9 MN840167 9-r b… 9-r ba… 9-р ба… 9-r bag… bag   bag        9 NA       1840167 
#> 10 MN840169 10-r … 10-r b… 10-р б… 10-r ba… bag   bag       10 NA       1840169 
#> # ℹ 81 more rows
#> # ℹ 6 more variables: parent_pcode <chr>, region_pcode <chr>,
#> #   aimag_pcode <chr>, soum_pcode <chr>, area_km2 <dbl>, has_geometry <lgl>

but not their boundaries. No open dataset publishes them: GADM, geoBoundaries, the humanitarian COD and OpenStreetMap all stop at soums. The National Statistics Office of Mongolia (NSO) shares bag boundaries on request.

Getting the file

Write to NSO at and ask for:

  • bag boundaries as a shapefile or GeoPackage;
  • each bag’s NSO code (7 digits, such as 1830151) or its soum code and bag number;
  • the date the boundaries are valid for;
  • the terms of use (can you share maps? can the file be redistributed?).

Using the file

Point mongolmaps at the file:

bags <- mn_bags(path = "bags_from_nso.gpkg")
mn_map(mn_bags(aimag = "Khovd", path = "bags_from_nso.gpkg"), label = TRUE)

or register it once, for example in your .Rprofile, and forget about it:

options(mongolmaps.bags_path = "C:/data/bags_from_nso.gpkg")

mn_bags(aimag = "Khovd")
mn_admin("bag") # khoroos and bags together
mn_join(my_bag_table, by = "code", level = "bag")

mn_read_bags() checks the file when it is first read:

  • it finds the column of bag codes (or use code_col);
  • every code must be a known NSO bag code;
  • every bag must lie inside its soum (turn off with check_nesting = FALSE).

Bags then get the same columns as every other map in the package, so joining NSO bag tables works like joining aimag tables.