Clean data, documented, ready to analyse.
Every dataset comes with its source, its date, its variable dictionary, its known limits and its licence. Three states, shown plainly: Available · Sponsorable · In preparation. We do not put a "Buy" button on a file that does not exist yet.
Year 1: Haiti only. Everything here concerns Haiti. Regional and African datasets will return in the second year, once the country is covered.
Looking for a complete solution for your organisation instead? See the decision packs and licences →
Collected, cleaned and documented by hand. One-time payment, instant download by card or PayPal, a written licence.
The free files are for learning. The paid ones are for delivering.
| Free | For sale | |
|---|---|---|
| What you open | A raw CSV | Several files already assembled, ready for Excel, Power BI and QGIS |
| Scope | One indicator, to practise on | Complete reference sets and series, up to 11,655 geolocated rows |
| Understanding the columns | A README file | Variable dictionary: every column, its unit, its source |
| Mapping | — | GeoJSON ready for Power BI and QGIS, official OCHA p-codes |
| The work already done | You clean and cross-reference it | Cross-checked sources, harmonised units, known limitations written down |
| Use | Learning, tutorials | Usage licence: thesis, client report, publication |
The maths is simple: cleaning and documenting one of these series by hand takes two to three days. That is what you are buying — not the numbers, which are often public, but the time.
570 communal sections + 140 chief towns + 10,945 populated places, all named in French and in Haitian Creole, with official OCHA p-codes, areas and WGS84 coordinates. Every populated place is attached to its communal section. GeoJSON for all three levels included, ready for Power BI and QGIS.
$45
See the datasheet & buyThe part of the Geo Pack covering a single department: sections, chief towns, populated places and polygons. For those working on one territory only.
$5
Buy — $5An Excel workbook ready to open: one sheet per commune chosen from a list, the table of the 140 communes, the detail with the status of every value, the age pyramid, the dictionary and the sources. The data stays free — what is sold is the assembly, and what it preserves: an absence stays empty, never filled with a zero.
$25 XLSX • 7 sheets • 1,251 absences left empty
Order the workbook — $25A Power BI project (PBIP) ready to open: 7 tables wired to the site's published data, relationships set on the p-codes, measures documented with their limitations. It refreshes in one click — the day Atmart updates a file, your reports follow. Built and validated in Power BI Desktop.
$50
Buy — $50A QGIS project ready to open: communes and departments, named watersheds, flood-prone zones, working and broken water points, filling stations and finance — styled remote layers, loaded from the site. Built and validated in QGIS.
$50
Buy — $50The part of the Geo Pack that concerns one department: communal sections, chief town, populated places and polygons. For anyone working on a single territory. The whole country remains the better deal — ten separate departments cost $90, the full pack $29.
$9 per department • one-off payment
Choose a departmentAn Excel workbook ready to open: one sheet per commune chosen from a list, the table of the 140 communes, the detail with the status of every value, the age pyramid, the dictionary and the sources. The data stays free — what is sold is the assembly, and what it preserves: an absence stays empty, never filled with a zero.
$25 XLSX • 7 sheets • 1,251 absences left empty
Order the workbook — $25A Power BI project (PBIP) ready to open: 7 tables wired to the site's published data, relationships set on the p-codes, measures documented with their limitations. It refreshes in one click — the day Atmart updates a file, your reports follow. Built and validated in Power BI Desktop.
$50
Buy — $50A QGIS project ready to open: communes and departments, named watersheds, flood-prone zones, working and broken water points, filling stations and finance — styled remote layers, loaded from the site. Built and validated in QGIS.
$50
Buy — $50Instant download, no account, no email. Every file comes with its documentation. CC BY 4.0 licence: cite the source and it is yours.
Remittances received in USD and as a % of GDP: Haiti, Dominican Republic, Jamaica, Senegal, Côte d'Ivoire, Mali, Cameroon, Togo. Real, documented data. Source: World Bank.
Free 26 rows • 6 columns
Download (CSV)Ten communal sections with their French and Haitian Creole names, OCHA p-codes, area and GPS coordinates: exactly the structure of the Geo Pack, so you can judge before buying.
Free 10 rows • 13 columns
Download (CSV)For every source we use: organisation, URL, licence, legal class, coverage, frequency and last verification date. What we are allowed to publish, and what we are not.
Free 35 rows • 17 columns
Download (CSV)The population of each commune by five-year bracket and by sex. This is what makes it possible to say how many school-age children or elderly people live in a territory — and no longer just how many inhabitants. These are projections, not a census: Haiti's last census dates from 2003.
$0 CSV • 7,140 rows • 18 columns
Download freeMonthly price readings, product by product and market by market: rice, beans, maize, oil, fuel. The network does not cover the whole country — 9 sentinel markets, one per department in nine departments out of ten. To be read as a signal, never as a national average.
$0 CSV • 14,140 readings • 32 products tracked
Download (CSV)The recorded markets: stable identifier, normalised name, attachment to the communal section and coordinates. This is the family whose coverage holds — 136 communes out of 140, and every record is geolocated.
$0 CSV + GeoJSON • 182 markets
Download freePublic spending actually executed at national level, for the current financial year, drawn from citable official documents. Two markers only, and that is deliberate: there is no breakdown by department or by commune, because the source does not publish one. Two sourced figures are worth more than an invented distribution.
Free 2 rows • 18 columns
Download freeWhich organisations work in which commune, and in which sector. These are presences declared by the organisations themselves to OCHA's 3W, not a field survey: an absence from this file does not prove an absence on the ground.
$0 CSV • 1,457 rows • 17 columns
Download freeNon-governmental organisations recognised by the State: registration number, date of recognition, field of work and declared areas. The contact details are institutional — those of the head office, not of individuals.
$0 CSV • 680 rows • 22 columns
Download freeHow many notaries and land surveyors are recorded in each commune. The count alone: no name, no address, no telephone. A zero means "none in our sources", never "none in the territory" — 55 communes out of 140 are covered.
$0 CSV • 280 rows • 9 columns
Download freeWhat each commune has in facilities, family by family. One row per commune and per type, with its source and its vintage — enough to compare territories without confusing what is counted with what is estimated.
$0 CSV • 1,899 rows • 8 columns
Download freeViolent events and deaths recorded, by commune and by month, from 2018 to today. A record of reported events: what is not reported does not appear.
Free 4,368 rows • 8 columns
Download (CSV)Displaced people recorded by IOM's Displacement Tracking Matrix. A dated field survey that does not claim to cover what the teams could not reach.
Free 150 rows • 9 columns
Download (CSV)The phases of the Integrated Food Security Phase Classification, by zone and period. An expert classification, not a direct measurement.
Free 210 rows • 12 columns
Download (CSV)Recorded water points and their functional status. Coverage is concentrated in the North and the Centre: elsewhere, no point does not mean no water.
Free 6,276 rows • 1 columns
Download (CSV)Filling stations mapped by OpenStreetMap contributors. Contributory mapping: complete nowhere, uneven from one commune to the next.
Free 310 rows
Download (CSV)Financial service points mapped by OpenStreetMap contributors. Same caveat: contributory, therefore uneven.
Free 474 rows • 1 columns
Download (CSV)The share of each commune's territory classified as flood-prone. Mind what this measures: an area mapped as flood-prone, not a probabilistic hazard by return period. The source's method is not documented.
Free 140 rows • 7 columns
Download (CSV)The extents mapped as flood-prone. Geometries simplified for display — not to be used for area calculation to the metre.
Free 134 rows
Download (CSV)The reference hydrological division, from the 2014 SRTM model. Geometries simplified for display.
Free 31 rows • 3 columns
Download (CSV)The breakdown of each commune's territory by broad land-cover class. The layer dates from 1998 — a quarter of a century: it serves to read a trend, not a current state. And it is an image classification, so what a sensor saw, not a cadastre.
Free 1,042 rows • 7 columns
Download (CSV)Communication towers and masts, mobile phone shops, operator offices and public internet access mapped in OpenStreetMap, coloured by operator (Digicel, Natcom, Voilà, Access Haiti). This is NOT the operators' network inventory: neither Digicel, nor Natcom, nor CONATEL publishes it. A commune without a point is not a commune without coverage.
Free 498 rows
Download (CSV)Count of telecom equipment by commune, by operator and by kind — 498 items across 81 communes out of 140. A contributory count: it measures the mapping effort as much as the real network.
Free 219 rows • 9 columns
Download (CSV)The population of each commune by five-year bracket and by sex. This is what makes it possible to say how many school-age children or elderly people live in a territory — and no longer just how many inhabitants. These are projections, not a census: Haiti's last census dates from 2003.
$0 CSV • 7,140 rows • 18 columns
Download freeMonthly price readings, product by product and market by market: rice, beans, maize, oil, fuel. The network does not cover the whole country — 9 sentinel markets, one per department in nine departments out of ten. To be read as a signal, never as a national average.
$0 CSV • 14,140 readings • 32 products tracked
Download freeThe recorded markets: stable identifier, normalised name, attachment to the communal section and coordinates. This is the family whose coverage holds — 136 communes out of 140, and every record is geolocated.
$0 CSV + GeoJSON • 182 markets
Download free GeoJSONPublic spending actually executed at national level, for the current financial year, drawn from citable official documents. Two markers only, and that is deliberate: there is no breakdown by department or by commune, because the source does not publish one. Two sourced figures are worth more than an invented distribution.
Free 2 rows • 18 columns
Download freeWhich organisations work in which commune, and in which sector. These are presences declared by the organisations themselves to OCHA's 3W, not a field survey: an absence from this file does not prove an absence on the ground.
$0 CSV • 1,457 rows • 17 columns
Download freeNon-governmental organisations recognised by the State: registration number, date of recognition, field of work and declared areas. The contact details are institutional — those of the head office, not of individuals.
$0 CSV • 680 rows • 22 columns
Download freeHow many notaries and land surveyors are recorded in each commune. The count alone: no name, no address, no telephone. A zero means "none in our sources", never "none in the territory" — 55 communes out of 140 are covered.
$0 CSV • 280 rows • 9 columns
Download freeWhat each commune has in facilities, family by family. One row per commune and per type, with its source and its vintage — enough to compare territories without confusing what is counted with what is estimated.
$0 CSV • 1,899 rows • 8 columns
Download freeSources identified, method written: what is missing is the funding for the preparation work. An institution can sponsor the opening of one of these datasets — it then becomes free for a whole cohort of students and researchers.
A standardised sheet per commune: territorial identity, population, area, localities, schools, health facilities, markets, roads, hazard exposure, ten socio-economic indicators — each with its source and quality score. The first product that uses the Backbone end to end.
6000 $ sponsorship
Sponsor this datasetA sentinel network of 9 markets is already published and downloadable, free of charge. This sponsorship would fund the next step: extending coverage to every market tracked by CNSA and FEWS NET, and keeping the series updated monthly.
2500 $ sponsorship
Sponsor this datasetGDP, inflation, gourde/USD exchange rate, imports and exports over 25 years, harmonised across sources.
2000 $ sponsorship
Sponsor this datasetEnrolment, institutions, results.
2500 $ sponsorship
Sponsor this datasetFacilities, staff, vaccination coverage.
3000 $ sponsorship
Sponsor this dataset