heimsmarkmid

# Heimsmarkmið — Iceland's UN Sustainable Development Goal statistics **Requires:** Tier 0 (core). Iceland's implementation of the UN 2030 Agenda. Hagstofa Íslands publishes 137 national SDG indicators (all 17 goals) via a public `open-sdg` instance. - **Site:** `heimsmarkmidin.hagstofa.is` (Icelandic) / `heimsmarkmidin.hagstofa.is/en/` (English) - **Data repo:** `hagstofan.github.io/heimsmarkmid-data-prod/` (GitHub Pages) - **Platform:** [open-sdg](https://open-sdg.org) — same toolkit used by UK, US, and ~20 other national SDG portals. If you know one, you know all of them. ## Architecture Frontend is a server-rendered Jekyll static site. Underlying data and metadata are published as flat files on a sibling GitHub Pages repo: | Path | Contents | |------|----------| | `/is/data/{code}.csv` | Time-series data with disaggregation columns | | `/is/meta/{code}.json` | Indicator metadata (name, source, graph config, units) | | `/is/headline/{code}.csv` | Headline series only (no disaggregation) | | `/is/comb/{code}.json` | Combined JSON (data + headline, structured for charting) | | `/is/zip/all_indicators.zip` | All 137 data CSVs in one 437 KB download | | `/en/...` | English-language mirror of the above | Indicator codes follow the UN SDG nomenclature: goal-target-indicator, e.g. `1-1-1` (Goal 1 · Target 1 · Indicator 1) or `16-b-1` (target "b", indicator 1). ## Usage ```bash # Download the ZIP, extract to data/raw/heimsmarkmid/, build catalog CSV uv run python scripts/heimsmarkmid.py fetch uv run python scripts/heimsmarkmid.py fetch --force # Re-download (ZIP is cached) # Print catalog uv run python scripts/heimsmarkmid.py list uv run python scripts/heimsmarkmid.py list --goal 4 # Fetch data + metadata for one indicator uv run python scripts/heimsmarkmid.py get 1-1-1 uv run python scripts/heimsmarkmid.py get 16-b-1 ``` The ZIP is fetched once and cached; metadata is fetched live on `get` (the open-sdg meta endpoint doesn't ship inside the ZIP). ## Catalog schema `data/processed/heimsmarkmid_catalog.csv`: | Column | Description | |--------|-------------| | `code` | Indicator code (e.g. `1-1-1`, `16-b-1`) | | `goal` | Leading SDG goal number (1–17) | | `target` | Middle segment (numeric or single letter) | | `indicator` | Final segment of the code | | `columns` | `\|`-separated disaggregation columns (excluding `Year`, `Value`) | | `n_rows` | Number of data rows in the CSV | | `first_year`, `last_year` | Range of years present (empty when no numeric rows) | ## Data-file schema Every `/is/data/{code}.csv` has: | Column | Meaning | |--------|---------| | `Year` | Calendar year (integer) | | `<disaggregations>` | Zero or more columns (`Kyn`, `Aldur`, `Svæði`, etc.); empty value = "Total" | | `Value` | Numeric measurement in the unit declared in the meta file | Unit is in `meta.computation_units` — common values are `Hundraðshlutar (%)`, `Fjöldi`, `krónur`. Units can vary within a single CSV if an indicator publishes multiple measurement series side-by-side (see `4-1-1` where the `Units` column differentiates reading vs. maths proficiency). ## Goal coverage (2026-04 snapshot) | Goal | Indicators | Total rows | |------|------------|-----------:| | 1 · No poverty | 9 | 856 | | 2 · Zero hunger | 7 | 124 | | 3 · Good health | 22 | 1,187 | | 4 · Quality education | 9 | 519 | | 5 · Gender equality | 7 | 491 | | 6 · Clean water | 8 | 48 | | 7 · Clean energy | 6 | 353 | | 8 · Decent work | 10 | 2,453 | | 9 · Industry/infrastructure | 6 | 1,458 | | 10 · Reduced inequalities | 5 | 267 | | 11 · Sustainable cities | 9 | 324 | | 12 · Responsible consumption | 8 | 595 | | 13 · Climate action | 2 | 958 | | 14 · Life below water | 2 | 6 | | 15 · Life on land | 7 | 94 | | 16 · Peace & institutions | 10 | 267 | | 17 · Partnerships | 10 | 338 | | **Total** | **137** | **~10,338** | ## Caveats 1. **Reporting gaps are normal.** Some indicators have `n_rows=0` — they're listed in the national framework but Iceland hasn't reported yet. Others have a single row (e.g. `4-a-1` has 2 rows, one per survey year). 2. **Disaggregation columns differ per indicator.** Don't assume a common schema across files — always join by `code` via the catalog, or inspect the header before parsing. 3. **"Total" rows use empty disaggregation cells.** When you need the headline series without breakdowns, filter rows where all disaggregation columns are empty — or use `/is/headline/{code}.csv` directly. 4. **Rows can duplicate across years with different units.** `4-1-1` reports both reading and maths proficiency in the same file, with a `Tegund færni` discriminator. Always disambiguate with the full key `(Year, *disaggregations)` before aggregating. 5. **Meta is English-labelled even on the `/is/` tree.** Field keys (`indicator_name`, `graph_title`) stay in English; only the values are translated. 6. **The data repo is rebuilt nightly.** If you're caching the ZIP, re-download with `fetch --force` when you need fresh values — the script does not compare timestamps for you. 7. **Overlap with [velsaeldarvisar](../velsaeldarvisar/SKILL.md).** About a dozen indicators (low-income share, gender gap, unemployment) appear in both publications but with different dimensional breakdowns. The SDG version follows UN methodology; the velsaeldarvísar version follows Eurostat/OECD. 8. **Encoding.** GitHub Pages serves all CSV/JSON/YAML as UTF-8. Icelandic chars appear in indicator titles, units (`Hundraðshlutar`, `krónur`), and disaggregation labels. Read with `encoding="utf-8"` and write JSON with `ensure_ascii=False`. ## Attribution "Byggir á upplýsingum frá Hagstofu Íslands (landsvísar fyrir heimsmarkmið)." ## Related - [velsaeldarvisar](../velsaeldarvisar/SKILL.md) — Hagstofa's wellbeing/social/cultural indicator portal. Separate content, partial thematic overlap. - [hagstofan](../hagstofan/SKILL.md) — many SDG indicators are re-computations over PX-Web tables reachable here.