Wikidata QID ↔︎ OCHA Pcode Crosswalk

Published

June 30, 2026

Overview

This document builds and evaluates a crosswalk between Wikidata QIDs and OCHA p-codes from the Global Administrative Boundaries GDB (maps/global_admin_boundaries_matched_latest.gdb), which contains:

  • ADM2: 19,705 features (provinces, departments, cantons, municipalities)
  • ADM3: 55,622 features (municipalities, districts, communes, parishes)

Six countries that appear in the Wikipedia presence analysis are absent from the GDB (Argentina, Brazil, Paraguay, Suriname, Uruguay, Cuba, Puerto Rico) and are excluded throughout. The 15 remaining countries in both datasets are the focus here.

Two strategies are applied in priority order:

  1. Code-based — for countries where a Wikidata external-ID property directly encodes the pcode suffix (e.g. P14142 for Bolivia, P7325 for Colombia). These are the most reliable matches and are preferred wherever available.
  2. Name-based — normalized string matching between Wikidata labels and GDB name columns, with accent and admin-type prefix stripping. For countries with many duplicate sub-unit names (Venezuela, Panama), a hierarchical second pass uses P131 (parent admin unit) to anchor items before matching.

The pipeline is implemented in src/pcode-crosswalk.R. QuickStatements to add P7590 (OCHA P-code) to matched items are generated at the end of this document but are not executed — a manual quality review is expected first.

Setup

Show code 
library(sf)
library(httr)
library(jsonlite)
library(dplyr)
library(purrr)
library(stringr)
library(tidyr)
library(tibble)
library(stringi)
library(ggplot2)

source("src/pcode-crosswalk.R")
source("R/create_quick_statement.R")

Data

GDB layers

Show code 
GDB_PATH <- "maps/global_admin_boundaries_matched_latest.gdb"

adm2_sf <- st_read(GDB_PATH, layer = "admin2", quiet = TRUE)
adm3_sf <- st_read(GDB_PATH, layer = "admin3", quiet = TRUE)

cat("ADM2:", nrow(adm2_sf), "features\n")
ADM2: 19705 features
Show code 
cat("ADM3:", nrow(adm3_sf), "features\n")
ADM3: 55622 features

Countries present in both datasets

The 15 countries present in both the GDB and the Wikipedia presence analysis, with their ISO 3166-1 alpha-2 codes used to join the two sources:

Show code 
# Countries in the Wikipedia presence analysis (from wikipedia-admin-presence.qmd)
# that also appear in the GDB
our_countries_gdb <- tribble(
  ~country,              ~country_qid, ~iso2,
  "Bolivia",             "Q750",       "BO",
  "Chile",               "Q298",       "CL",
  "Colombia",            "Q739",       "CO",
  "Costa Rica",          "Q800",       "CR",
  "Dominican Republic",  "Q786",       "DO",
  "Ecuador",             "Q736",       "EC",
  "El Salvador",         "Q792",       "SV",
  "Guatemala",           "Q774",       "GT",
  "Haiti",               "Q790",       "HT",
  "Honduras",            "Q783",       "HN",
  "Mexico",              "Q96",        "MX",
  "Nicaragua",           "Q811",       "NI",
  "Panama",              "Q804",       "PA",
  "Peru",                "Q419",       "PE",
  "Venezuela",           "Q717",       "VE"
)

knitr::kable(
  our_countries_gdb |> select(country, iso2),
  caption = "Countries present in both the GDB and the Wikipedia presence analysis"
)
Countries present in both the GDB and the Wikipedia presence analysis
country iso2
Bolivia BO
Chile CL
Colombia CO
Costa Rica CR
Dominican Republic DO
Ecuador EC
El Salvador SV
Guatemala GT
Haiti HT
Honduras HN
Mexico MX
Nicaragua NI
Panama PA
Peru PE
Venezuela VE

GDB working tables

Strip geometry, attach country labels, and select the columns used for matching and name lookups.

Show code 
adm2_gdb <- adm2_sf |>
  st_drop_geometry() |>
  inner_join(our_countries_gdb |> select(country, iso2), by = "iso2") |>
  select(country, iso2, adm0_pcode, adm1_pcode, adm2_pcode,
         adm2_name, adm2_name1, adm2_name2, adm2_name3)

adm3_gdb <- adm3_sf |>
  st_drop_geometry() |>
  inner_join(our_countries_gdb |> select(country, iso2), by = "iso2") |>
  select(country, iso2, adm1_pcode, adm2_pcode, adm3_pcode,
         adm3_name, adm3_name1, adm3_name2, adm3_name3)

cat("ADM2 rows in scope:", nrow(adm2_gdb), "\n")
ADM2 rows in scope: 5673
Show code 
cat("ADM3 rows in scope:", nrow(adm3_gdb), "\n")
ADM3 rows in scope: 11178

Wikidata items

all_data is loaded from the cached presence-matrix pipeline in wikipedia-admin-presence.qmd. It provides one row per Wikidata item with its English and Spanish labels and admin level.

Show code 
# Load presence data from the admin presence pipeline cache
sa_adm2_all   <- readRDS("data/cache/sa_adm2_all.rds")
sa_adm3_all   <- readRDS("data/cache/sa_adm3_all.rds")
newc_adm2_all <- readRDS("data/cache/newc_adm2_all.rds")
newc_adm3_all <- readRDS("data/cache/newc_adm3_all.rds")

brazil_cache <- "data/cache/sa_adm2_brazil_df.rds"
all_tiers <- list(sa_adm2_all, sa_adm3_all, newc_adm2_all, newc_adm3_all)
if (file.exists(brazil_cache)) all_tiers <- c(all_tiers, list(readRDS(brazil_cache)))

all_data <- bind_rows(all_tiers)

# Keep only the columns used by the crosswalk pipeline
all_data <- all_data |>
  select(qid, country, adm_level, label_en, label_es) |>
  filter(country %in% our_countries_gdb$country)

cat("Wikidata items in scope:", nrow(all_data), "\n")
Wikidata items in scope: 11657
Show code 
cat("Countries:", n_distinct(all_data$country), "\n")
Countries: 15
Show code 
all_data |> count(adm_level)
# A tibble: 2 × 2
  adm_level     n
  <chr>     <int>
1 ADM2       5738
2 ADM3       5919

Code-based property reference

The following table records which Wikidata property encodes the pcode suffix for each country. The pcode is reconstructed as paste0(iso2, property_value). Property coverage was verified by sampling items and comparing constructed codes against the GDB pcode lists.

Show code 
knitr::kable(
  PCODE_CODE_PROPERTIES |>
    mutate(adm_levels = map_chr(adm_levels, paste, collapse = " + ")) |>
    select(Country = country, Property = pid, `ISO2 prefix` = iso2,
           `Admin levels` = adm_levels),
  caption = "Wikidata properties used for code-based pcode matching"
)
Wikidata properties used for code-based pcode matching
Country Property ISO2 prefix Admin levels
Bolivia P14142 BO c(“ADM2”, “ADM3”)
Colombia P7325 CO c(“ADM2”, “ADM3”)
Mexico P3801 MX ADM2
Costa Rica P281 CR ADM3
Chile P6929 CL ADM3
Haiti P1370 HT ADM2

Properties investigated but not adopted:

Country Property Reason not used
Argentina P4833 (INDEC) Argentina absent from GDB
Colombia P2082 Not populated on sampled items; P7325 preferred
Haiti P1370 Very sparse — most arrondissements lack it
All P7590 (OCHA P-code) Not yet populated — this crosswalk will add it

Name-matching methodology

For countries without a reliable code property, labels are normalized and compared to GDB name columns:

  1. Normalize: lowercase → Latin-ASCII transliteration (strip accents) → collapse non-alphanumeric characters to spaces → trim.
  2. Strip admin-type prefix: remove leading tokens such as “Cantón”, “Provincia de”, “Municipio de” etc. (both languages, both normalized and stripped variants are compared).
  3. Match within country: inner join on (country, normalized_name). Only one-to-one matches are kept (n_gdb_hits == 1).
  4. Hierarchical pass (Venezuela, Panama): for ADM3 items with ambiguous country-level matches, fetch the P131 parent via SPARQL, look the parent up in the ADM2 crosswalk, and re-match within the confirmed parent unit.

Build the crosswalk

Show code 
crosswalk <- build_pcode_crosswalk(
  all_data               = all_data,
  adm2_gdb               = adm2_gdb,
  adm3_gdb               = adm3_gdb,
  hierarchical_countries = c("Venezuela", "Panama"),
  sparql_batch_size      = 300,
  sparql_delay           = 0.4,
  cache_file             = "data/cache/crosswalk_qid_pcode.rds"
)

cat("Total matched items:", nrow(crosswalk), "\n")
Total matched items: 9742
Show code 
cat("Unique QIDs:        ", n_distinct(crosswalk$qid), "\n")
Unique QIDs:         9742

Coverage summary

Show code 
gdb_countries <- our_countries_gdb$country

total_in_scope <- all_data |>
  count(country, adm_level, name = "n_wikidata")

coverage <- crosswalk |>
  count(country, adm_level, match_method, name = "n_matched") |>
  left_join(total_in_scope, by = c("country", "adm_level")) |>
  mutate(pct = round(100 * n_matched / n_wikidata, 1))

# Roll up to country × adm_level for a clean summary
coverage_summary <- crosswalk |>
  count(country, adm_level, name = "n_matched") |>
  left_join(total_in_scope, by = c("country", "adm_level")) |>
  mutate(
    pct            = round(100 * n_matched / n_wikidata, 1),
    primary_method = case_when(
      country == "Bolivia"    ~ "code (P14142)",
      country == "Colombia"   ~ "code (P7325)",
      country == "Mexico"     ~ "code (P3801)",
      country == "Costa Rica" & adm_level == "ADM3" ~ "code (P281)",
      country == "Chile"      & adm_level == "ADM3" ~ "code (P6929)",
      TRUE ~ "name match"
    )
  ) |>
  arrange(country, adm_level)

knitr::kable(
  coverage_summary |>
    select(Country = country, Level = adm_level,
           `WD items` = n_wikidata, Matched = n_matched,
           `%` = pct, `Primary method` = primary_method),
  caption = "Crosswalk coverage by country and administrative level"
)
Crosswalk coverage by country and administrative level
Country Level WD items Matched % Primary method
Bolivia ADM2 112 112 100.0 code (P14142)
Bolivia ADM3 341 339 99.4 code (P14142)
Chile ADM2 57 53 93.0 name match
Chile ADM3 485 353 72.8 code (P6929)
Colombia ADM2 1107 1104 99.7 code (P7325)
Colombia ADM3 108 3 2.8 code (P7325)
Costa Rica ADM2 84 84 100.0 name match
Costa Rica ADM3 495 492 99.4 code (P281)
Ecuador ADM2 222 203 91.4 name match
Ecuador ADM3 568 340 59.9 name match
El Salvador ADM2 45 44 97.8 name match
Guatemala ADM2 341 318 93.3 name match
Haiti ADM2 43 28 65.1 name match
Honduras ADM2 298 240 80.5 name match
Mexico ADM2 2481 2461 99.2 code (P3801)
Nicaragua ADM2 153 149 97.4 name match
Panama ADM2 81 72 88.9 name match
Panama ADM3 698 583 83.5 name match
Peru ADM2 205 192 93.7 name match
Peru ADM3 1892 1589 84.0 name match
Venezuela ADM2 352 216 61.4 name match
Venezuela ADM3 1186 767 64.7 name match
Show code 
coverage_summary |>
  mutate(
    strategy = if_else(str_starts(primary_method, "code"), "Code-based", "Name-based"),
    country  = reorder(country, pct, FUN = mean)
  ) |>
  ggplot(aes(x = pct, y = country, fill = strategy)) +
  geom_col(position = position_dodge2(preserve = "single", padding = 0.15)) +
  geom_vline(xintercept = 100, linetype = "dashed", linewidth = 0.4) +
  facet_wrap(~ adm_level) +
  scale_fill_manual(values = c("Code-based" = "#2166AC", "Name-based" = "#F4A582")) +
  scale_x_continuous(limits = c(0, 105), breaks = seq(0, 100, 25)) +
  labs(x = "% of Wikidata items matched", y = NULL, fill = "Strategy") +
  theme_minimal(base_size = 12) +
  theme(legend.position = "bottom")
Figure 1: Crosswalk coverage (%) by country and admin level, coloured by match strategy

Unmatched items

Show code 
unmatched <- all_data |>
  anti_join(crosswalk, by = "qid") |>
  count(country, adm_level, name = "n_unmatched") |>
  left_join(total_in_scope, by = c("country", "adm_level")) |>
  mutate(pct_unmatched = round(100 * n_unmatched / n_wikidata, 1)) |>
  filter(n_unmatched > 0) |>
  arrange(desc(n_unmatched))

knitr::kable(
  unmatched |>
    select(Country = country, Level = adm_level,
           Unmatched = n_unmatched, `Total WD` = n_wikidata,
           `% unmatched` = pct_unmatched),
  caption = "Unmatched Wikidata items by country and level"
)
Unmatched Wikidata items by country and level
Country Level Unmatched Total WD % unmatched
Venezuela ADM3 419 1186 35.3
Peru ADM3 303 1892 16.0
Ecuador ADM3 228 568 40.1
Dominican Republic ADM2 157 157 100.0
Haiti ADM3 146 146 100.0
Venezuela ADM2 136 352 38.6
Chile ADM3 132 485 27.2
Panama ADM3 115 698 16.5
Colombia ADM3 105 108 97.2
Honduras ADM2 58 298 19.5
Guatemala ADM2 23 341 6.7
Mexico ADM2 20 2481 0.8
Ecuador ADM2 19 222 8.6
Haiti ADM2 15 43 34.9
Peru ADM2 13 205 6.3
Panama ADM2 9 81 11.1
Chile ADM2 4 57 7.0
Nicaragua ADM2 4 153 2.6
Colombia ADM2 3 1107 0.3
Costa Rica ADM3 3 495 0.6
Bolivia ADM3 2 341 0.6
El Salvador ADM2 1 45 2.2

The largest gaps are:

  • Venezuela ADM2/ADM3: municipality names repeat heavily across states; a full state-level hierarchical pass via P131 would recover many of the remaining ~400 unmatched items.
  • Ecuador ADM3: many parish names are non-unique across cantons and Wikidata coverage of parishes is incomplete.
  • Chile ADM3: P6929 covers ~58%; the remainder have no P6929 value in Wikidata.
  • Panama ADM3: corregimiento names repeat within districts.

Quality check

Match method breakdown

Show code 
crosswalk |>
  count(match_method, name = "n") |>
  mutate(
    pct      = round(100 * n / nrow(crosswalk), 1),
    strategy = if_else(str_starts(match_method, "code"), "Code-based", "Name-based")
  ) |>
  arrange(strategy, match_method) |>
  knitr::kable(caption = "Crosswalk entries by match method")
Crosswalk entries by match method
match_method n pct strategy
code_P14142 451 4.6 Code-based
code_P281 491 5.0 Code-based
code_P3801 2457 25.2 Code-based
code_P6929 281 2.9 Code-based
code_P7325 1100 11.3 Code-based
name_adm2 1607 16.5 Name-based
name_adm3 3145 32.3 Name-based
name_adm3_hierarchical 210 2.2 Name-based

Spot-check: name-matched items

The function below pulls a stratified sample of name-matched items and shows the Wikidata label alongside the GDB name it matched. Scan for false positives (same normalized string, different real-world unit).

Show code 
spot_check_name_matches <- function(crosswalk, all_data, adm2_gdb, adm3_gdb,
                                    country_filter = NULL, n = 20, seed = 5501) {
  set.seed(seed)

  xw <- crosswalk |>
    filter(str_starts(match_method, "name")) |>
    left_join(all_data |> select(qid, label_en, label_es), by = "qid")

  if (!is.null(country_filter)) xw <- xw |> filter(country %in% country_filter)

  adm2_names <- adm2_gdb |> distinct(adm2_pcode, gdb_name = adm2_name)
  adm3_names <- adm3_gdb |> distinct(adm3_pcode, gdb_name = adm3_name)

  xw |>
    left_join(adm2_names, by = "adm2_pcode") |>
    left_join(adm3_names, by = "adm3_pcode") |>
    mutate(gdb_name = coalesce(gdb_name.x, gdb_name.y)) |>
    select(qid, country, adm_level, match_method,
           wd_label_en = label_en, wd_label_es = label_es,
           gdb_name, pcode) |>
    slice_sample(n = min(n, nrow(xw)))
}
Show code 
spot_check_name_matches(crosswalk, all_data, adm2_gdb, adm3_gdb, n = 30) |>
  knitr::kable(caption = "Random sample of name-matched items for review")
Random sample of name-matched items for review
qid country adm_level match_method wd_label_en wd_label_es gdb_name pcode
Q3313291 Peru ADM3 name_adm3 San Vicente de Cañete District Distrito de San Vicente de Cañete San Vicente de Cañete PE150501
Q611291 Venezuela ADM3 name_adm3 Parroquia Lezama Parroquia Lezama Lezama VE120902
Q4848913 Peru ADM3 name_adm3 Bajo Biavo District Distrito de Bajo Biavo Bajo Biavo PE220203
Q21157181 Panama ADM3 name_adm3 Sambú Sambú Sambú PA050112
Q2393454 Honduras ADM2 name_adm2 Magdalena Magdalena Magdalena HN1008
Q5139413 Peru ADM3 name_adm3 Cochorco District Distrito de Cochorco Cochorco PE130903
Q16639801 Venezuela ADM3 name_adm3 Granados Granados Granados VE210303
Q28225642 Panama ADM3 name_adm3 Barrio Colón NA Barrio Colón PA110404
Q606975 Venezuela ADM2 name_adm2 Municipio Aragua Municipio Aragua Aragua VE0302
Q3313347 Peru ADM3 name_adm3 Santiago de Anchucaya District Distrito de Santiago de Anchucaya Santiago de Anchucaya PE150729
Q2393269 Honduras ADM2 name_adm2 Camasca Camasca Camasca HN1002
Q6087669 Panama ADM3 name_adm3 Progreso Progreso Progreso PA020203
Q3274 Nicaragua ADM2 name_adm2 Managua Managua Managua NI5525
Q568999 Guatemala ADM2 name_adm2 San Luis Jilotepeque San Luis Jilotepeque San Luis Jilotepeque GT2103
Q3050086 Venezuela ADM3 name_adm3 El Molino El Molino El Molino VE140504
Q2402824 Guatemala ADM2 name_adm2 Cuyotenango Cuyotenango Cuyotenango GT1002
Q3471094 Venezuela ADM3 name_adm3 NA Samuel Darío Maldonado Samuel Darío Maldonado VE202201
Q6319436 Peru ADM3 name_adm3 Ninabamba District Distrito de Ninabamba Ninabamba PE061306
Q111543533 Ecuador ADM3 name_adm3 Cumandá Cumandá Cumanda EC061050
Q2113237 Honduras ADM2 name_adm2 San Lorenzo San Lorenzo San Lorenzo HN1709
Q5047837 Peru ADM3 name_adm3 Casa Grande District Distrito de Casa Grande Casa Grande PE130208
Q23660288 Chile ADM3 name_adm3 Linares Linares Linares CL07401
Q2216549 Panama ADM2 name_adm2 Renacimiento District Renacimiento Renacimiento PA0210
Q16681069 Venezuela ADM3 name_adm3 Tres Esquinas Tres Esquinas Tres Esquinas VE211807
Q1729098 Honduras ADM2 name_adm2 Yuscarán Yuscarán Yuscaran HN0701
Q2405941 Costa Rica ADM2 name_adm2 Santo Domingo Canton Cantón de Santo Domingo Santo Domingo CR403
Q1958679 Honduras ADM2 name_adm2 Danlí Danlí Danli HN0703
Q2702986 Peru ADM3 name_adm3 Uranmarca District Distrito de Uranmarca Uranmarca PE030607
Q1023655 Guatemala ADM2 name_adm2 Salamá Salamá Salamá GT1501
Q2697794 Peru ADM3 name_adm3 Llacllin District Distrito de Llacllín Llacllin PE021705

Code-matched sample

Code matches are structurally reliable (property value = numeric pcode suffix), but it is worth confirming a few against the GDB name to catch any systematic encoding errors.

Show code 
set.seed(1783)

adm2_names <- adm2_gdb |> distinct(adm2_pcode, gdb_name_adm2 = adm2_name)
adm3_names <- adm3_gdb |> distinct(adm3_pcode, gdb_name_adm3 = adm3_name)

crosswalk |>
  filter(str_starts(match_method, "code")) |>
  left_join(all_data |> select(qid, label_en), by = "qid") |>
  left_join(adm2_names, by = "adm2_pcode") |>
  left_join(adm3_names, by = "adm3_pcode") |>
  mutate(gdb_name = coalesce(gdb_name_adm2, gdb_name_adm3)) |>
  select(qid, country, adm_level, pcode, match_method, label_en, gdb_name) |>
  group_by(match_method) |>
  slice_sample(n = 5) |>
  ungroup() |>
  knitr::kable(caption = "Sample of code-matched items (5 per property)")
Sample of code-matched items (5 per property)
qid country adm_level pcode match_method label_en gdb_name
Q1303284 Bolivia ADM3 BO021402 code_P14142 Coripata Municipality Coripata
Q491527 Bolivia ADM3 BO020902 code_P14142 Sapahaqui Municipality Sapahaqui
Q1552096 Bolivia ADM3 BO020905 code_P14142 Cairoma Municipality Cairoma
Q1107940 Bolivia ADM3 BO070104 code_P14142 La Guardia La Guardia
Q647771 Bolivia ADM3 BO040503 code_P14142 Cruz De Machacamarca Cruz de Machacamarca
Q3946914 Costa Rica ADM3 CR10308 code_P281 San Cristóbal San Cristobal
Q2033984 Costa Rica ADM3 CR61301 code_P281 Puerto Jiménez Puerto Jiménez
Q3677436 Costa Rica ADM3 CR30704 code_P281 Cipreses Cipreses
Q3211297 Costa Rica ADM3 CR11401 code_P281 San Vicente San Vicente
Q3941341 Costa Rica ADM3 CR20607 code_P281 El Rosario El Rosario
Q16488640 Mexico ADM2 MX24003 code_P3801 Aquismón Municipality Aquismón
Q20275700 Mexico ADM2 MX16027 code_P3801 Chucándiro Municipality Chucándiro
Q2502727 Mexico ADM2 MX08040 code_P3801 Madera Municipality Madera
Q20290867 Mexico ADM2 MX16070 code_P3801 Purépero Municipality Purépero
Q3294614 Mexico ADM2 MX16077 code_P3801 San Lucas Municipality San Lucas
Q13063 Chile ADM3 CL09106 code_P6929 Galvarino Galvarino
Q200965 Chile ADM3 CL08312 code_P6929 Tucapel Tucapel
Q3735 Chile ADM3 CL04202 code_P6929 Canela Canela
Q13049 Chile ADM3 CL09105 code_P6929 Freire Freire
Q13048 Chile ADM3 CL08104 code_P6929 Florida Florida
Q1525701 Colombia ADM2 CO25799 code_P7325 Tenjo Tenjo
Q444511 Colombia ADM2 CO13030 code_P7325 Altos del Rosario Altos del Rosario
Q2434056 Colombia ADM2 CO13620 code_P7325 San Cristóbal San Cristóbal
Q634563 Colombia ADM2 CO66075 code_P7325 Balboa Balboa
Q1442348 Colombia ADM2 CO25817 code_P7325 Tocancipá Tocancipá

Hierarchical ADM3 matches

These 210 items were ambiguous at the country level but resolved via P131. A higher proportion of these merit individual review.

Show code 
set.seed(2290)

crosswalk |>
  filter(match_method == "name_adm3_hierarchical") |>
  left_join(all_data |> select(qid, label_en, label_es), by = "qid") |>
  left_join(adm3_names, by = "adm3_pcode") |>
  select(qid, country, pcode, adm2_pcode, wd_label_en = label_en,
         wd_label_es = label_es, gdb_name = gdb_name_adm3) |>
  slice_sample(n = 20) |>
  knitr::kable(caption = "Sample of hierarchically resolved ADM3 matches")
Sample of hierarchically resolved ADM3 matches
qid country pcode adm2_pcode wd_label_en wd_label_es gdb_name
Q21014463 Panama PA130205 PA1302 El Potrero El Potrero El Potrero
Q21065781 Panama PA070205 PA0702 Leones Leones Leones
Q6115163 Panama PA030607 PA0306 Río Grande Río Grande Río Grande
Q5489478 Venezuela VE231302 VE2313 Parroquia Bolívar Parroquia Bolívar Bolívar
Q16675617 Venezuela VE210505 VE2105 Santa Cruz Santa Cruz Santa Cruz
Q16302718 Panama PA040505 PA0405 Palmira Palmira Palmira
Q21042281 Panama PA130703 PA1307 Leones Leones Leones
Q16675509 Venezuela VE210407 VE2104 San José San José San José
Q16651832 Venezuela VE211402 VE2114 NA La Concepción La Concepción
Q6059076 Panama PA020406 PA0204 Palmira Palmira Palmira
Q16301879 Panama PA110503 PA1105 Guayabito Guayabito Guayabito
Q20018155 Panama PA090221 PA0902 Santo Domingo Santo Domingo Santo Domingo
Q3009627 Venezuela VE200501 VE2005 NA Cárdenas Cárdenas
Q3472574 Venezuela VE151203 VE1512 Santa Barbara Santa Bárbara Santa Barbará
Q5822769 Panama PA070604 PA0706 El Pedregoso El Pedregoso El Pedregoso
Q3472641 Venezuela VE060404 VE0604 Parroquia Santa Inés Parroquia Santa Inés Santa Inés
Q3471642 Venezuela VE160811 VE1608 Parroquia San Vicente Parroquia San Vicente San Vicente
Q20018151 Panama PA090219 PA0902 San José San José San José
Q3049698 Venezuela VE090602 VE0906 Parroquia El Amparo Parroquia El Amparo El Amparo
Q16303982 Venezuela VE210802 VE2108 Arnoldo Gabaldon Arnoldo Gabaldón Arnoldo Gabaldon

Items with duplicate GDB names (potential false positives)

Name matches where the GDB contains more than one unit with the same normalized name in the same country are by construction excluded from the crosswalk (n_gdb_hits == 1 filter). This section documents how many items were dropped for this reason per country.

Show code 
# Re-run the name join without the n_gdb_hits filter to count ambiguous cases
normalize_admin_name <- function(x) {
  x |> str_to_lower() |> stri_trans_general("Latin-ASCII") |>
    str_replace_all("[^a-z0-9]+", " ") |> str_squish()
}

wd_adm2_labels <- all_data |>
  filter(adm_level == "ADM2") |>
  pivot_longer(c(label_en, label_es), values_to = "raw", names_to = "lang") |>
  filter(!is.na(raw)) |>
  mutate(norm = normalize_admin_name(raw)) |>
  distinct(qid, country, norm)

gdb_adm2_norm <- adm2_gdb |>
  pivot_longer(starts_with("adm2_name"), values_to = "raw", names_to = "nv") |>
  filter(!is.na(raw), raw != "") |>
  mutate(norm = normalize_admin_name(raw)) |>
  distinct(country, adm2_pcode, norm)

wd_adm2_labels |>
  inner_join(gdb_adm2_norm, by = c("country", "norm"),
             relationship = "many-to-many") |>
  group_by(qid, country) |>
  summarise(n_gdb_hits = n_distinct(adm2_pcode), .groups = "drop") |>
  filter(n_gdb_hits > 1) |>
  count(country, name = "n_ambiguous") |>
  arrange(desc(n_ambiguous)) |>
  knitr::kable(caption = "ADM2 items dropped due to ambiguous name match (>1 GDB unit)")
ADM2 items dropped due to ambiguous name match (>1 GDB unit)
country n_ambiguous
Colombia 148
Honduras 46
Venezuela 16
Guatemala 12
Mexico 12

Unmatched entities — high-coverage countries

For country × level combinations with ≥ 85% match rate the residual gap is small enough to inspect unit by unit. Both sides of the join are shown: Wikidata items with no pcode, and GDB units with no QID.

Show code 
high_coverage <- coverage_summary |>
  filter(pct >= 85) |>
  select(country, adm_level, n_matched, n_wikidata, pct)

knitr::kable(
  high_coverage |>
    select(Country = country, Level = adm_level,
           Matched = n_matched, `WD total` = n_wikidata, `%` = pct),
  caption = "Country × level strata with ≥ 85% match rate"
)
Country × level strata with ≥ 85% match rate
Country Level Matched WD total %
Bolivia ADM2 112 112 100.0
Bolivia ADM3 339 341 99.4
Chile ADM2 53 57 93.0
Colombia ADM2 1104 1107 99.7
Costa Rica ADM2 84 84 100.0
Costa Rica ADM3 492 495 99.4
Ecuador ADM2 203 222 91.4
El Salvador ADM2 44 45 97.8
Guatemala ADM2 318 341 93.3
Mexico ADM2 2461 2481 99.2
Nicaragua ADM2 149 153 97.4
Panama ADM2 72 81 88.9
Peru ADM2 192 205 93.7

Unmatched Wikidata items

These items exist in Wikidata but have no GDB pcode in the crosswalk. Likely causes in high-coverage countries: the unit was created, renamed, or split after the GDB snapshot; the Wikidata label is in an unexpected language or spelling; or the Wikidata item covers a slightly different geographic entity (e.g. a city item vs. a municipality item).

Show code 
unmatched_wd <- all_data |>
  semi_join(high_coverage, by = c("country", "adm_level")) |>
  anti_join(crosswalk,  by = "qid") |>
  mutate(
    wd_link = paste0("[", qid, "](https://www.wikidata.org/wiki/", qid, ")")
  ) |>
  arrange(country, adm_level, label_en) |>
  select(Country = country, Level = adm_level,
         QID = wd_link, `Label (en)` = label_en, `Label (es)` = label_es)

cat(nrow(unmatched_wd), "unmatched Wikidata items across",
    n_distinct(paste(unmatched_wd$Country, unmatched_wd$Level)), "strata\n")
101 unmatched Wikidata items across 11 strata
Show code 
knitr::kable(unmatched_wd,
             caption = "Wikidata items with no GDB pcode (high-coverage strata)")
Wikidata items with no GDB pcode (high-coverage strata)
Country Level QID Label (en) Label (es)
Bolivia ADM3 Q1324520 El Puente Municipality NA
Bolivia ADM3 Q106409795 San Pedro de Macha Municipality NA
Chile ADM2 Q201131 Bío Bío province Biobío
Chile ADM2 Q201277 Coyhaique Province Coihaique
Chile ADM2 Q721611 San Felipe de Aconcagua Province San Felipe de Aconcagua
Chile ADM2 Q721755 Ñuble Province Provincia de Ñuble
Colombia ADM2 Q1526223 Belén de Bajirá Belén de Bajirá
Colombia ADM2 Q10315074 La Paz (Cesar) La Paz (Cesar)
Colombia ADM2 Q136413477 San Rafael (Antioquia) San Rafael (Antioquia)
Costa Rica ADM3 Q136387746 Cabagra Cabagra
Costa Rica ADM3 Q135095023 Conte Burica Conte Burica
Costa Rica ADM3 Q134516942 Pijije Pijije
Ecuador ADM2 Q3655923 Baños Canton Cantón Baños
Ecuador ADM2 Q1992892 Bolívar Canton Cantón Bolívar
Ecuador ADM2 Q2450110 Bolívar Canton Cantón Bolívar
Ecuador ADM2 Q1989893 Coronel Marcelino Maridueña Canton Cantón Coronel Marcelino Maridueña
Ecuador ADM2 Q607595 El Empalme Canton Cantón El Empalme
Ecuador ADM2 Q1992935 Francisco de Orellana Canton Cantón Francisco de Orellana
Ecuador ADM2 Q1990118 General Antonio Elizalde Canton Cantón General Antonio Elizalde
Ecuador ADM2 Q1990553 Logroño Canton Cantón Logroño
Ecuador ADM2 Q502238 Olmedo Canton Cantón Olmedo
Ecuador ADM2 Q1989972 Olmedo Canton Cantón Olmedo
Ecuador ADM2 Q3656220 Pelileo Canton Cantón Pelileo
Ecuador ADM2 Q3656240 Píllaro Canton Cantón Píllaro
Ecuador ADM2 Q2466472 Quito Metro Distrito Metropolitano de Quito
Ecuador ADM2 Q1990505 Rumiñahui Canton Cantón Rumiñahui
Ecuador ADM2 Q2707803 Salcedo Canton Cantón San Miguel de Salcedo
Ecuador ADM2 Q1444202 Santiago de Méndez Canton Cantón Santiago de Méndez
Ecuador ADM2 Q131718210 Sevilla Don Bosco Canton Cantón Sevilla Don Bosco
Ecuador ADM2 Q1992944 Veinticuatro de Mayo Canton Cantón Veinticuatro de mayo
Ecuador ADM2 Q1990029 Yaguachi Canton Cantón Yaguachi
El Salvador ADM2 Q94339265 San Sebastián Analco San Sebastián Analco
Guatemala ADM2 Q1555 Guatemala City Ciudad de Guatemala
Guatemala ADM2 Q129369 La Democracia La Democracia
Guatemala ADM2 Q2608785 La Democracia La Democracia
Guatemala ADM2 Q164481 La Libertad La Libertad
Guatemala ADM2 Q1748790 La Libertad La Libertad
Guatemala ADM2 Q21279786 Petapán Petapán
Guatemala ADM2 Q7396574 Sacatepéquez Sacatepéquez
Guatemala ADM2 Q2401814 San Ildefonso Ixtahuacán San Ildefonso Ixtahuacán
Guatemala ADM2 Q741169 San José San José
Guatemala ADM2 Q3471500 San José San José
Guatemala ADM2 Q1254234 San Juan Comalapa San Juan Comalapa
Guatemala ADM2 Q2718157 San Juan Ostuncalco San Juan Ostuncalco
Guatemala ADM2 Q845976 San Lorenzo San Lorenzo
Guatemala ADM2 Q3947493 San Lorenzo San Lorenzo
Guatemala ADM2 Q921966 San Miguel Petapa San Miguel Petapa
Guatemala ADM2 Q317444 San Pedro Sacatepéquez San Pedro Sacatepéquez
Guatemala ADM2 Q2608367 San Pedro Sacatepéquez San Pedro Sacatepéquez
Guatemala ADM2 Q2402459 San Pedro Soloma San Pedro Soloma
Guatemala ADM2 Q1748767 Santa Bárbara Santa Bárbara
Guatemala ADM2 Q2401903 Santa Bárbara Santa Bárbara
Guatemala ADM2 Q2390990 Santa Cruz Barillas Santa Cruz Barillas
Guatemala ADM2 Q1727528 Santa Cruz El Chol Santa Cruz el Chol
Guatemala ADM2 Q332147 Santa María Cahabón Santa María Cahabón
Mexico ADM2 Q14623618 Belisario Domínguez Municipality Belisario Domínguez
Mexico ADM2 Q51120215 Capitán Luis Ángel Vidal Municipality Municipio de Capitán Luis Ángel Vidal
Mexico ADM2 Q112326765 Coatetelco Municipality Municipio de Coatetelco
Mexico ADM2 Q65173865 Dzitbalché Municipality Municipio de Dzitbalché
Mexico ADM2 Q14630098 El Parral Municipality Municipio de El Parral
Mexico ADM2 Q120691691 Eldorado Municipality Municipio de Eldorado
Mexico ADM2 Q14630115 Emiliano Zapata Municipality Municipio de Emiliano Zapata
Mexico ADM2 Q102426425 Honduras de la Sierra Municipality Municipio de Honduras de la Sierra
Mexico ADM2 Q120692070 Juan José Ríos Municipality Municipio de Juan José Ríos
Mexico ADM2 Q108382175 Las Vigas Municipality Municipio de Las Vigas
Mexico ADM2 Q14902858 Mezcalapa Municipality Municipio de Mezcalapa
Mexico ADM2 Q21451666 Puerto Morelos Municipality Municipio de Puerto Morelos
Mexico ADM2 Q51120325 Rincón Chamula San Pedro Municipality Municipio de Rincón Chamula San Pedro
Mexico ADM2 Q107774002 San Felipe Municipality Municipio de San Felipe
Mexico ADM2 Q108384850 San Nicolás Municipality Municipio de San Nicolás
Mexico ADM2 Q13912475 San Quintín Municipality Municipio de San Quintín
Mexico ADM2 Q108385750 Santa Cruz del Rincón Municipality Municipio de Santa Cruz del Rincón
Mexico ADM2 Q61959987 Seybaplaya Municipality Municipio de Seybaplaya
Mexico ADM2 Q127726695 Villa de Pozos Municipality Municipio de Villa de Pozos
Mexico ADM2 Q108382060 Ñuu Savi Municipality Municipio de Ñuu Savi
Nicaragua ADM2 Q2394975 El Jícaro El Jícaro
Nicaragua ADM2 Q1647889 Kukra Hill Kukra Hill
Nicaragua ADM2 Q1647867 San Juan de Cinco Pinos San Juan de Cinco Pinos
Nicaragua ADM2 Q2217458 Waspán Waspán
Panama ADM2 Q2074220 Almirante Almirante
Panama ADM2 Q28663520 Gorgona Distrito de Gorgona
Panama ADM2 Q17635339 Jirondai District Jirondai
Panama ADM2 Q751010 Kusapín District ó SABORIKÄTE Saborikäte
Panama ADM2 Q49685881 Omar Torrijos Herrera Distrito Especial Omar Torrijos Herrera
Panama ADM2 Q17629454 Santa Catalina o Calovébora District Santa Catalina o Calovébora
Panama ADM2 Q3711451 Santa María District Santa María (distrito de Herrera)
Panama ADM2 Q21065819 Tierras Altas District Distrito de Tierras Altas
Panama ADM2 Q3711636 Ñürüm District Ñürün
Peru ADM2 Q1166121 Antonio Raymondi Province Antonio Raimondi
Peru ADM2 Q4790285 Arica Arica
Peru ADM2 Q16622605 Chancay Province provincia de Chancay
Peru ADM2 Q9063356 Conchucos Province provincia de Conchucos
Peru ADM2 Q2634400 Constitutional Province of Callao Provincia Constitucional del Callao
Peru ADM2 Q731515 Huallaga Province Provincia del Huallaga
Peru ADM2 Q9063361 Iquique provincia de Iquique
Peru ADM2 Q6088838 Moquegua Province provincia de Moquegua
Peru ADM2 Q1806437 Paucar del Sara Sara Province Páucar del Sarasara
Peru ADM2 Q5240570 Tarapacá Tarapacá
Peru ADM2 Q7685484 Tarapacá Tarapacá
Peru ADM2 Q84575363 Tinta province Provincia de Tinta
Peru ADM2 Q537708 Vilcas Huamán Province Vilcashuamán

Unmatched GDB units

These GDB units have no matched Wikidata item. Likely causes: the unit has no Wikidata item yet (a gap in Wikidata coverage); the GDB reflects a more recent or more granular boundary dataset than what Wikidata tracks; or the pcode belongs to a sub-division that Wikidata models at a different level.

Show code 
# ADM2 GDB units with no crosswalk entry, for high-coverage countries
unmatched_gdb_adm2 <- adm2_gdb |>
  semi_join(high_coverage |> filter(adm_level == "ADM2"), by = "country") |>
  filter(!adm2_pcode %in% crosswalk$pcode) |>
  mutate(adm_level = "ADM2") |>
  select(Country = country, Level = adm_level, Pcode = adm2_pcode,
         `ADM1 pcode` = adm1_pcode, `GDB name` = adm2_name)

# ADM3 GDB units with no crosswalk entry, for high-coverage countries
unmatched_gdb_adm3 <- adm3_gdb |>
  semi_join(high_coverage |> filter(adm_level == "ADM3"), by = "country") |>
  filter(!adm3_pcode %in% crosswalk$pcode) |>
  mutate(adm_level = "ADM3") |>
  select(Country = country, Level = adm_level, Pcode = adm3_pcode,
         `ADM1 pcode` = adm1_pcode, `GDB name` = adm3_name)

unmatched_gdb <- bind_rows(unmatched_gdb_adm2, unmatched_gdb_adm3) |>
  arrange(Country, Level, Pcode)

cat(nrow(unmatched_gdb), "unmatched GDB units across",
    n_distinct(paste(unmatched_gdb$Country, unmatched_gdb$Level)), "strata\n")
84 unmatched GDB units across 8 strata
Show code 
knitr::kable(unmatched_gdb,
             caption = "GDB units with no matched Wikidata QID (high-coverage strata)")
GDB units with no matched Wikidata QID (high-coverage strata)
Country Level Pcode ADM1 pcode GDB name
Chile ADM2 CL057 CL05 San Felipe
Chile ADM2 CL083 CL08 Bío-Bío
Chile ADM2 CL111 CL11 Coyhaique
Colombia ADM2 CO91263 CO91 El Encanto
Colombia ADM2 CO91405 CO91 La Chorrera
Colombia ADM2 CO91407 CO91 La Pedrera
Colombia ADM2 CO91430 CO91 La Victoria
Colombia ADM2 CO91460 CO91 Mirití - Paraná
Colombia ADM2 CO91530 CO91 Puerto Alegría
Colombia ADM2 CO91536 CO91 Puerto Arica
Colombia ADM2 CO91669 CO91 Puerto Santander
Colombia ADM2 CO91798 CO91 Tarapacá
Colombia ADM2 CO94663 CO94 Mapiripana
Colombia ADM2 CO94883 CO94 San Felipe
Colombia ADM2 CO94884 CO94 Puerto Colombia
Colombia ADM2 CO94885 CO94 La Guadalupe
Colombia ADM2 CO94886 CO94 Cacahual
Colombia ADM2 CO94887 CO94 Pana Pana
Colombia ADM2 CO94888 CO94 Morichal
Colombia ADM2 CO97511 CO97 Pacoa
Colombia ADM2 CO97777 CO97 Papunahua
Colombia ADM2 CO97889 CO97 Yavaraté
Ecuador ADM2 EC0402 EC04 Bolivar
Ecuador ADM2 EC0505 EC05 Salcedo
Ecuador ADM2 EC0908 EC09 Empalme
Ecuador ADM2 EC0920 EC09 San Jacinto de Yaguachi
Ecuador ADM2 EC0923 EC09 Crnel. Marcelino Maridueña
Ecuador ADM2 EC0927 EC09 Gnral. Antonio Elizalde
Ecuador ADM2 EC1116 EC11 Olmedo
Ecuador ADM2 EC1302 EC13 Bolivar
Ecuador ADM2 EC1316 EC13 24 de Mayo
Ecuador ADM2 EC1318 EC13 Olmedo
Ecuador ADM2 EC1405 EC14 Santiago
Ecuador ADM2 EC1410 EC14 Logroðo
Ecuador ADM2 EC1701 EC17 Quito
Ecuador ADM2 EC1705 EC17 Rumiðahui
Ecuador ADM2 EC1802 EC18 Baños de Agua Santa
Ecuador ADM2 EC1807 EC18 San Pedro de Pelileo
Ecuador ADM2 EC1808 EC18 Santiago de Pillaro
Ecuador ADM2 EC2201 EC22 Orellana
Ecuador ADM2 EC9001 EC90 Las Golondrinas
Ecuador ADM2 EC9003 EC90 Manga del Cura
Ecuador ADM2 EC9004 EC90 El Piedrero
El Salvador ADM2 SV03999 SV03 Embalse Cerron Grande
El Salvador ADM2 SV10899 SV10 Lago de Llopango
El Salvador ADM2 SV12998 SV12 Lago de Guija
El Salvador ADM2 SV12999 SV12 Lago de Coatepeque
Guatemala ADM2 GT0100 GT01 Lago De Amatitlan
Guatemala ADM2 GT0101 GT01 Guatemala
Guatemala ADM2 GT0109 GT01 San Pedro Sacatepéquez
Guatemala ADM2 GT0117 GT01 Petapa
Guatemala ADM2 GT0404 GT04 Comalapa
Guatemala ADM2 GT0503 GT05 La Democracia
Guatemala ADM2 GT0509 GT05 San José
Guatemala ADM2 GT0700 GT07 Lago De Atitlan
Guatemala ADM2 GT0909 GT09 Ostuncalco
Guatemala ADM2 GT1007 GT10 San Lorenzo
Guatemala ADM2 GT1015 GT10 Santa Bárbara
Guatemala ADM2 GT1202 GT12 San Pedro Sacatepéquez
Guatemala ADM2 GT1229 GT12 San Lorenzo
Guatemala ADM2 GT1308 GT13 Soloma
Guatemala ADM2 GT1309 GT13 Ixtahuacán
Guatemala ADM2 GT1310 GT13 Santa Bárbara
Guatemala ADM2 GT1311 GT13 La Libertad
Guatemala ADM2 GT1312 GT13 La Democracia
Guatemala ADM2 GT1326 GT13 Barillas
Guatemala ADM2 GT1333 GT13 Petatán
Guatemala ADM2 GT1506 GT15 El Chol
Guatemala ADM2 GT1612 GT16 Cahabón
Guatemala ADM2 GT1702 GT17 San José
Guatemala ADM2 GT1703 GT17 San Benito
Guatemala ADM2 GT1705 GT17 La Libertad
Nicaragua ADM2 NI0515 NI05 Jícaro
Nicaragua ADM2 NI3015 NI30 Cinco Pinos
Nicaragua ADM2 NI9107 NI91 Waspam
Nicaragua ADM2 NI9330 NI93 Kukrahill
Panama ADM2 PA0707 PA07 Santa María
Panama ADM2 PA0801 PA08 Kuna Yala
Panama ADM2 PA1005 PA10 Ñürüm
Panama ADM2 PA1007 PA10 Kusapín
Peru ADM2 PE0203 PE02 Antonio Raymondi
Peru ADM2 PE0508 PE05 Paucar del Sara Sara
Peru ADM2 PE0511 PE05 Vilcas Huaman
Peru ADM2 PE2204 PE22 Huallaga

Summary: gap by country

Show code 
wd_gap <- unmatched_wd |>
  count(Country, Level, name = "unmatched_wd")

gdb_gap <- unmatched_gdb |>
  count(Country, Level, name = "unmatched_gdb")

high_coverage |>
  left_join(wd_gap,  by = c("country" = "Country", "adm_level" = "Level")) |>
  left_join(gdb_gap, by = c("country" = "Country", "adm_level" = "Level")) |>
  replace_na(list(unmatched_wd = 0L, unmatched_gdb = 0L)) |>
  mutate(
    `Unmatched WD`  = unmatched_wd,
    `Unmatched GDB` = unmatched_gdb
  ) |>
  select(Country = country, Level = adm_level, `%` = pct,
         `Unmatched WD`, `Unmatched GDB`) |>
  knitr::kable(caption = "Residual gap on each side for high-coverage strata")
Residual gap on each side for high-coverage strata
Country Level % Unmatched WD Unmatched GDB
Bolivia ADM2 100.0 0 0
Bolivia ADM3 99.4 2 0
Chile ADM2 93.0 4 3
Colombia ADM2 99.7 3 19
Costa Rica ADM2 100.0 0 0
Costa Rica ADM3 99.4 3 0
Ecuador ADM2 91.4 19 21
El Salvador ADM2 97.8 1 4
Guatemala ADM2 93.3 23 25
Mexico ADM2 99.2 20 0
Nicaragua ADM2 97.4 4 4
Panama ADM2 88.9 9 4
Peru ADM2 93.7 13 4

QuickStatements

The function make_pcode_quickstatements() generates Wikidata QuickStatements to add P7590 (OCHA P-code) to matched items. It takes the crosswalk, a country name, and optional reference metadata, and returns a tibble with one row per item.

No statements are executed here. Run the quality checks above first, then call the function with appropriate methods and reference_* arguments for each country.

Function signature

make_pcode_quickstatements(
  crosswalk,
  country,                        # e.g. "Bolivia"
  pid            = "P7590",       # OCHA P-code property
  adm_levels     = c("ADM2","ADM3"),
  methods        = NULL,          # NULL = all; or e.g. c("code_P14142")
  reference_qid  = NULL,          # Wikidata item for "stated in" (P248)
  reference_url  = NULL,          # URL for P854
  retrieved_date = Sys.Date(),
  output_file    = NULL           # write TSV if provided
)

Preview: Bolivia (code-matched items only)

Bolivia is used as the preview country because all its matches are code-based (P14142) and therefore require no additional name-match review.

Show code 
qs_bolivia <- make_pcode_quickstatements(
  crosswalk     = crosswalk,
  country       = "Bolivia",
  methods       = c("code_P14142"),       # code-confirmed only
  reference_url = "https://data.humdata.org/dataset/global-administrative-unit-layers"
)

cat("Statements to generate:", nrow(qs_bolivia), "\n\n")
Statements to generate: 451
Show code 
cat("Sample (first 5):\n")
Sample (first 5):
Show code 
cat(head(qs_bolivia$quick_statement, 5), sep = "\n")
Q1001757 | P7590 | "BO070401" | S854 | "https://data.humdata.org/dataset/global-administrative-unit-layers" | S813 | +2026-06-28T00:00:00Z/11
Q1024796 | P7590 | "BO070703" | S854 | "https://data.humdata.org/dataset/global-administrative-unit-layers" | S813 | +2026-06-28T00:00:00Z/11
Q1029614 | P7590 | "BO070706" | S854 | "https://data.humdata.org/dataset/global-administrative-unit-layers" | S813 | +2026-06-28T00:00:00Z/11
Q1107940 | P7590 | "BO070104" | S854 | "https://data.humdata.org/dataset/global-administrative-unit-layers" | S813 | +2026-06-28T00:00:00Z/11
Q1107950 | P7590 | "BO070704" | S854 | "https://data.humdata.org/dataset/global-administrative-unit-layers" | S813 | +2026-06-28T00:00:00Z/11
Show code 
qs_bolivia |>
  select(qid, adm_level, pcode, match_method, quick_statement) |>
  head(10) |>
  knitr::kable(caption = "First 10 QuickStatements for Bolivia (not yet uploaded)")
First 10 QuickStatements for Bolivia (not yet uploaded)
qid adm_level pcode match_method quick_statement
Q1001757 ADM3 BO070401 code_P14142 Q1001757 | P7590 | “BO070401” | S854 | “https://data.humdata.org/dataset/global-administrative-unit-layers” | S813 | +2026-06-28T00:00:00Z/11
Q1024796 ADM3 BO070703 code_P14142 Q1024796 | P7590 | “BO070703” | S854 | “https://data.humdata.org/dataset/global-administrative-unit-layers” | S813 | +2026-06-28T00:00:00Z/11
Q1029614 ADM3 BO070706 code_P14142 Q1029614 | P7590 | “BO070706” | S854 | “https://data.humdata.org/dataset/global-administrative-unit-layers” | S813 | +2026-06-28T00:00:00Z/11
Q1107940 ADM3 BO070104 code_P14142 Q1107940 | P7590 | “BO070104” | S854 | “https://data.humdata.org/dataset/global-administrative-unit-layers” | S813 | +2026-06-28T00:00:00Z/11
Q1107950 ADM3 BO070704 code_P14142 Q1107950 | P7590 | “BO070704” | S854 | “https://data.humdata.org/dataset/global-administrative-unit-layers” | S813 | +2026-06-28T00:00:00Z/11
Q1107963 ADM3 BO070803 code_P14142 Q1107963 | P7590 | “BO070803” | S854 | “https://data.humdata.org/dataset/global-administrative-unit-layers” | S813 | +2026-06-28T00:00:00Z/11
Q1107968 ADM3 BO071201 code_P14142 Q1107968 | P7590 | “BO071201” | S854 | “https://data.humdata.org/dataset/global-administrative-unit-layers” | S813 | +2026-06-28T00:00:00Z/11
Q1107974 ADM3 BO070101 code_P14142 Q1107974 | P7590 | “BO070101” | S854 | “https://data.humdata.org/dataset/global-administrative-unit-layers” | S813 | +2026-06-28T00:00:00Z/11
Q1108148 ADM3 BO071301 code_P14142 Q1108148 | P7590 | “BO071301” | S854 | “https://data.humdata.org/dataset/global-administrative-unit-layers” | S813 | +2026-06-28T00:00:00Z/11
Q1108155 ADM3 BO070901 code_P14142 Q1108155 | P7590 | “BO070901” | S854 | “https://data.humdata.org/dataset/global-administrative-unit-layers” | S813 | +2026-06-28T00:00:00Z/11

Generating statements for other countries

Once the quality review is complete for a given country, write its statements to a file and review in the QuickStatements batch tool before submitting:

# Example — do not run until quality review is done
make_pcode_quickstatements(
  crosswalk     = crosswalk,
  country       = "Mexico",
  methods       = "code_P3801",
  reference_url = "https://data.humdata.org/dataset/global-administrative-unit-layers",
  output_file   = "data/qs_mexico_pcode.tsv"
)
Country Recommended methods before upload Notes
Bolivia code_P14142 100% coverage, no review needed
Colombia code_P7325 99% coverage, spot-check recommended
Mexico code_P3801 99% ADM2 coverage
Costa Rica ADM3 code_P281 99% ADM3; ADM2 by name needs review
Chile ADM3 code_P6929 58% coverage; name matches need review
All others name_adm2, name_adm3 Review spot-check table above first

Session info

Show code 
sessionInfo()
R version 4.5.2 (2025-10-31)
Platform: aarch64-apple-darwin20
Running under: macOS Sequoia 15.7.7

Matrix products: default
BLAS:   /System/Library/Frameworks/Accelerate.framework/Versions/A/Frameworks/vecLib.framework/Versions/A/libBLAS.dylib 
LAPACK: /Library/Frameworks/R.framework/Versions/4.5-arm64/Resources/lib/libRlapack.dylib;  LAPACK version 3.12.1

locale:
[1] en_US.UTF-8/en_US.UTF-8/en_US.UTF-8/C/en_US.UTF-8/en_US.UTF-8

time zone: America/Denver
tzcode source: internal

attached base packages:
[1] stats     graphics  grDevices utils     datasets  methods   base     

other attached packages:
 [1] ggplot2_4.0.3  stringi_1.8.7  tibble_3.3.1   tidyr_1.3.2    stringr_1.6.0 
 [6] purrr_1.2.2    dplyr_1.2.1    jsonlite_2.0.0 httr_1.4.7     sf_1.1-1      

loaded via a namespace (and not attached):
 [1] gtable_0.3.6       compiler_4.5.2     tidyselect_1.2.1   Rcpp_1.1.1-1.1    
 [5] dichromat_2.0-0.1  scales_1.4.0       yaml_2.3.12        fastmap_1.2.0     
 [9] R6_2.6.1           generics_0.1.4     classInt_0.4-11    knitr_1.51        
[13] htmlwidgets_1.6.4  units_0.8-7        DBI_1.2.3          RColorBrewer_1.1-3
[17] pillar_1.11.1      rlang_1.2.0        xfun_0.59          S7_0.2.2          
[21] otel_0.2.0         cli_3.6.6          withr_3.0.3        magrittr_2.0.5    
[25] class_7.3-23       digest_0.6.39      grid_4.5.2         rstudioapi_0.19.0 
[29] lifecycle_1.0.5    vctrs_0.7.3        KernSmooth_2.23-26 proxy_0.4-27      
[33] evaluate_1.0.5     glue_1.8.1         farver_2.1.2       e1071_1.7-16      
[37] rmarkdown_2.31     tools_4.5.2        pkgconfig_2.0.3    htmltools_0.5.9