English | Français | العربية

@geoalgeria/pharmacies

Algeria's pharmacies — as data you can install.

Overview

3,790 pharmacies (officines) across 67 wilayas, from OpenStreetMap — every record geocoded, bilingual (FR/AR) where named, with phone, opening hours and a dispensing flag where OSM has them, plus wilaya/commune linkage.

Installation

npm install @geoalgeria/pharmacies

Quick start

import {
  pharmacies,
  pharmacyById,
  pharmaciesByWilaya,
  metadata,
} from "@geoalgeria/pharmacies";

pharmacies().length; // 3790

// Everything in Alger (code 16)
pharmaciesByWilaya(16).length;

// A single record, and dataset metadata
pharmacyById("16-00001")?.commune;
metadata().wilayas_covered; // 67

What you can build

• A pharmacy locator / map layer for Algeria
• A "pharmacie de garde" base layer (join with opening hours)
• Health-access analysis by wilaya/commune

What's inside

| Dataset | Count | Coordinates | Notes | | --- | --- | --- | --- | | pharmacies | 3,790 | all 3,790 | OpenStreetMap, geocoded, wilaya/commune-linked |

• Named: 2,459 · with phone: 146 · with opening hours: 255 · with address: 1,159 · with dispensing: 524

Formats

• data/pharmacies.json — full array (typed by types/index.d.ts)
• data/csv/pharmacies.csv — flat CSV
• data/geojson/pharmacies.geojson — FeatureCollection (all records)
• data/metadata.json — counts, sources, generated date

import data from "@geoalgeria/pharmacies/data/pharmacies.json" with { type: "json" };

import type { Pharmacy } from "@geoalgeria/pharmacies";

How the data is built

Extracted from OpenStreetMap via Overpass (amenity=pharmacy across Algeria), de-duplicated (same-name-within-40 m and coincident points), then linked to wilaya/commune by nearest-centroid join against the geoalgeria commune set. Names are routed strictly by script so name_ar is always Arabic and name_fr always Latin, even when OSM mis-tags them. A bulk-import artifact — 1,769 unnamed, consecutive-id pharmacy ways dumped into one small area near Attatba (Tipaza) — is detected and excluded (it is not real officines). Rebuild with npm run fetch (or --cache to rebuild from the cached pull). See research/pharmacies/ in the monorepo.

On accuracy & coverage

Coverage is partial. 3,790 pharmacies are mapped in OpenStreetMap against an estimated ~11,000 officines nationally (an order-of-magnitude reference — there is no open official registry; the Ordre National des Pharmaciens portal is down). Coverage is uneven by wilaya and denser in the north — this is a community-maintained extract, not an official registry.

Coordinates are OSM node points (surveyed) or building-outline centroids (geo_precision). Commune is a nearest-centroid best-effort; wilaya is effectively exact. Names, phones and hours are only present where an OSM contributor tagged them.

Source & license

Data © OpenStreetMap contributors, licensed ODbL 1.0. Any distribution of this data (or derived DB) must credit OpenStreetMap and keep it under ODbL. Wilaya/commune linkage uses the geoalgeria base dataset. Package code under MIT (see LICENSE).

Questions?

Open an issue: https://github.com/yasserstudio/geoalgeria/issues