# Indonesia — HS 2106 imports, 2023 — verified data pack

**What this is:** one dataset you can check, not an opinion you have to trust.

## What is inside
`Indonesia_HS2106_Trade_Data_2023.xlsx` — 5 sheets:

| Sheet | Contents |
|---|---|
| Read me | scope, method, how to verify |
| Imports by partner | 5 partner countries, value (USD) and share (%), with `name_source` |
| Total | Indonesia HS 2106 imports 2023 = **USD 703,645,643** |
| World Bank indicators | GDP, population, internet penetration — each with year and source |
| Sources and gaps | every source URL, plus **what is deliberately missing** |

## How every figure is sourced
- Trade figures: **UN Comtrade** public preview API — reporter `360` (Indonesia), flow `M` (imports), `cmdCode=2106`, `period=2023`.
- Total is the Comtrade `partnerCode=0` (World) row. We verified it equals the sum of the 87 single-partner rows to the dollar.
- Partner names: **mapped by us from UN M49 codes**. The Comtrade payload returned no partner names, so every row states this in `name_source`. If you prefer code-only, ignore the name column.
- Macro figures: **World Bank** indicators, each with its own year.

## What is deliberately NOT here
- **Year-on-year change**: Comtrade returned HTTP 429 for 2022. We show it as *not available* instead of estimating it.
- **Sub-segment breakouts**: HS 2106 does not split "plant-based ready meals" from other 2106 products. Anything finer would be a guess.
- **Prices, margins, forecasts**: not in the source data, therefore not in the file.

## Verify it in 5 minutes
1. Open the Comtrade URL in `Sources and gaps`.
2. Find the row with `partnerCode = 0`. Its value should equal **703,645,643**.
3. Compare any partner row (e.g. `156` → 156,669,628 → 22.27%).
4. Check the World Bank series against the indicator code in the same sheet.

If any number does not match, tell us and we will correct it or refund.

## Use
You may use this data in your own reports, decks and client work. No attribution required. Do not resell the file as-is.

## Contact
Reply to the order message with the country + HS code you need next; typical turnaround for a new pack is 48 hours.


## Optional add-on: analysis note (English)
`Analysis_note_EN.md` — a 16,000-character English briefing built strictly on this dataset.

How it was checked (`lint_result.txt`, 9 of 9 checks passed):
- no invented firm names or signatures (a-no-fabricated-claims)
- no unexplained company names (b)
- every money amount and percentage has a source label within 200 characters (c)
- 76.6% of its numbers are traceable verbatim to the facts file (d)
- English ratio 0.9983 (e)
- the one unavailable fact (YoY) is explicitly disclosed as not verified (f)
- every country name exists in the data and carries a `name_source` (g)
- the top-5 shares sum to 66.28% and each equals value / total (h)
- the total is not any single partner row (i)

Two editorial labels were added by us during review, and they say so in the text:
"(shares computed from UN Comtrade 2023 partner values)" and
"(computed as 100% minus the top-5 shares, from UN Comtrade 2023)".
No figure was changed.