Add README.md

README.md

@@ -5,7 +5,7 @@ language_creators:
 - found
 language:
 - en
-license: other
+license: cc-by-sa-4.0
 multilinguality:
 - monolingual
 size_categories:
@@ -14,69 +14,152 @@ source_datasets:
 - original
 task_categories:
 - tabular-classification
+- other
 task_ids: []
 tags:
 - africa
 - humanitarian
 - hdx
 - electric-sheep-africa
+- affected-population
 - demographics
-- health
+- flooding
+- hazards-and-risk
+- health-facilities
+- indicators
 - ken
-pretty_name: Kenya - National Demographic and Health Data
+pretty_name: "Kenya - Risk Assessment Indicators"
 dataset_info:
-  features:
-  - name: adm2_pcode
-    dtype: string
-  - name: adm_pcode
-    dtype: string
-  - name: female_pop
-    dtype: int64
-  - name: children_u5
-    dtype: int64
-  - name: female_u5
-    dtype: int64
-  - name: elderly
-    dtype: int64
-  - name: pop_u15
-    dtype: int64
-  - name: female_u15
-    dtype: int64
-  - name: esa_source
-    dtype: string
-  - name: esa_processed
-    dtype: string
   splits:
-  - name: train
-    num_bytes: 21576
-    num_examples: 232
-  - name: test
-    num_bytes: 5394
-    num_examples: 58
-  download_size: 21943
-  dataset_size: 26970
-configs:
-- config_name: default
-  data_files:
-  - split: train
-    path: data/train-*
-  - split: test
-    path: data/test-*
+    - name: train
+      num_examples: 232
+    - name: test
+      num_examples: 58
 ---
 
-# Kenya - National Demographic and Health Data
+# Kenya - Risk Assessment Indicators
 
-**Publisher:** The DHS Program · **Source:** [HDX](https://data.humdata.org/dataset/dhs-data-for-kenya) · **License:** `hdx-other` · **Updated:** 2026-04-20
+**Publisher:** HeiGIT (Heidelberg Institute for Geoinformation Technology) · **Source:** [HDX](https://data.humdata.org/dataset/kenya---risk-assessment-indicators) · **License:** `cc-by-sa` · **Updated:** 2026-04-13
 
 ---
 
 ## Abstract
 
-Contains data from the [DHS data portal](https://api.dhsprogram.com/). There is also a dataset containing [Kenya - Subnational Demographic and Health Data](https://data.humdata.org/dataset/dhs-subnational-data-for-kenya) on HDX.
+This dataset provides comprehensive **Risk Assessment Indicators** for **Kenya**, aggregated at **admin level 2** and 
+can in particular be used to perform a structured risk assessment for **flood** hazards.
+It includes demographic, environmental, infrastructure, accessibility, and hazard-related data to support disaster risk and resilience analysis.
 
-The DHS Program Application Programming Interface (API) provides software developers access to aggregated indicator data from The Demographic and Health Surveys (DHS) Program. The API can be used to create various applications to help analyze, visualize, explore and disseminate data on population, health, HIV, and nutrition from more than 90 countries.
+All layers are derived from [HeiGIT’s GAIA Pipeline](https://giscience.github.io/gis-training-resource-center/content/GIS_AA/en_gaia_indicators_processing.html), integrating open data sources such as [WorldPop](https://www.worldpop.org/), 
+[OpenStreetMap](https://www.openstreetmap.org/), and [Google Earth Engine](https://earthengine.google.com/) based on 
+[HDX COD-AB](https://data.humdata.org/dataset/?q=cod-ab) boundaries.
 
-Each row in this dataset represents country-level aggregates. Data was last updated on HDX on 2026-04-20. Geographic scope: **KEN**.
+---
+
+### **Data Overview**
+
+- **Access to Services (`KEN_ADM2_access`)**  
+- **Facilities (`KEN_ADM2_facilities`)**  
+- **Coping Capacity (`KEN_ADM2_coping`)**  
+- **Demographics (`KEN_ADM2_demographics`)**  
+- **Rural Population (`KEN_ADM2_rural_population`)**  
+- **Vulnerability (`KEN_ADM2_vulnerability`)**  
+- **Flood Exposure (`KEN_ADM2_flood_exposure`)**  
+
+
+<p>&nbsp;</p>
+<p>&nbsp;</p>
+
+---
+
+### **Indicator Descriptions**
+
+#### **Access to Services (`KEN_ADM2_access`)**
+Represents the share of the population with access to key facilities within defined distances or travel times.
+
+- **ADM2_PCODE** – Administrative division code (ADM2)
+- **access_pop_education_5km / 10km / 20km** – Population within 5, 10, and 20 km of educational facilities
+- **access_pop_hospitals_30min / 1h / 2h** – Population within 30 minutes, 1 hour, and 2 hours of a hospital
+- **access_pop_primary_healthcare_30min / 1h / 2h** – Population within 30 minutes, 1 hour, and 2 hours of a primary health care facility
+
+Data Source: [openrouteservice (ORS)](https://openrouteservice.org/)
+
+---
+
+#### **Facilities (`KEN_ADM2_facilities`)**
+Counts of essential service facilities within each district.
+
+- **ADM2_PCODE** – Administrative division code (ADM2)
+- **education_count** – Number of educational facilities
+- **hospitals_count** – Number of hospitals
+- **primary_healthcare_count** – Number of primary health care facilities
+
+Data Source: [OpenStreetMap (OSM)](https://www.openstreetmap.org)
+
+---
+
+#### **Coping Capacity (`KEN_ADM2_coping`)**
+Combines **Access to Services** and **Facilities** data to represent a district’s coping capacity.
+
+---
+
+#### **Demographics (`KEN_ADM2_demographics`)**
+Shows the population composition by age and gender.
+
+- **ADM2_PCODE** – Administrative division code (ADM2)
+- **female_pop** – Total female population
+- **children_u5** – Population under 5 years old
+- **female_u5** – Female population under 5 years old
+- **elderly** – Population aged 65 and older
+- **pop_u15** – Population under 15 years old
+- **female_u15** – Female population under 15 years old
+
+Data Source: [Worldpop](https://www.worldpop.org/)
+
+---
+
+#### **Rural Population (`KEN_ADM2_rural_population`)**
+Same demographic breakdown as above, but limited to rural populations. Rural areas are those outside urban extents,
+typically characterized by lower population density, agricultural or natural land use, and limited infrastructure compared to urban centers.
+
+- **ADM2_PCODE** – Administrative division code (ADM2)
+- **female_pop_rural**, **children_u5_rural**, **female_u5_rural**, **elderly_rural**, **pop_u15_rural**, **female_u15_rural** – Rural demographic counts
+- **rural_pop_perc** – Percentage of total population living in rural areas
+
+Data Source: [Global Human Settlement Layer (GHSL)](https://human-settlement.emergency.copernicus.eu/datasets.php)
+
+---
+
+#### **Vulnerability (`KEN_ADM2_vulnerability`)**
+Combines **Demographics** and **Rural Population** indicators.
+
+---
+
+#### **Flood Exposure (`KEN_ADM2_flood_exposure`)**
+Shows population and facility exposure to flooding at 30 cm depth for multiple return periods.
+
+- **ADM2_PCODE** – Administrative division code (ADM2)
+- **female_pop_30cm**, **children_u5_30cm**, **female_u5_30cm**, **elderly_30cm**, **pop_u15_30cm**, **female_u15_30cm** – Exposed population by group
+- **education_30cm_pct / count**, **hospitals_30cm_pct / count**, **primary_healthcare_30cm_pct / count** – Facility exposure (percentage and count)
+
+Data Source: [The Joint Research Centre (JRC)](https://data.jrc.ec.europa.eu/collection/id-0054)
+
+
+---
+
+
+
+### **QGIS Plugin Risk Assessment Inputs**
+
+- **Coping Capacity** = Access + Facilities  
+- **Vulnerability** = Demographics + Rural Population  
+- **Exposure** = Vulnerable Population + Facilities exposed to Floods
+
+This dataset is part of HeiGIT’s **Risk Assessment Indicator Collection** on HDX.
+See more at [HeiGIT on HDX](https://data.humdata.org/organization/heidelberg-institute-for-geoinformation-technology) and learn about HeiGIT’s research at [HeiGIT](https://heigit.org/).  
+
+We are happy to hear about your use-cases — contact us at [communications@heigit.org](mailto:communications@heigit.org)!
+
+Each row in this dataset represents tabular records. Data was last updated on HDX on 2026-04-13. Geographic scope: **KEN**.
 
 *Curated into ML-ready Parquet format by [Electric Sheep Africa](https://huggingface.co/electricsheepafrica).*
 
@@ -87,26 +170,26 @@ Each row in this dataset represents country-level aggregates. Data was last upda
 | | |
 |---|---|
 | **Domain** | Public health |
-| **Unit of observation** | Country-level aggregates |
-| **Rows (total)** | 191 |
-| **Columns** | 27 (12 numeric, 15 categorical, 0 datetime) |
-| **Train split** | 152 rows |
-| **Test split** | 38 rows |
+| **Unit of observation** | Tabular records |
+| **Rows (total)** | 290 |
+| **Columns** | 10 (6 numeric, 4 categorical, 0 datetime) |
+| **Train split** | 232 rows |
+| **Test split** | 58 rows |
 | **Geographic scope** | KEN |
-| **Publisher** | The DHS Program |
-| **HDX last updated** | 2026-04-20 |
+| **Publisher** | HeiGIT (Heidelberg Institute for Geoinformation Technology) |
+| **HDX last updated** | 2026-04-13 |
 
 ---
 
 ## Variables
 
-**Geographic** — `iso3` (KEN), `dhs_countrycode` (KE), `countryname` (Kenya), `surveyyear` (range 1989.0–2022.0), `surveyid` (KE2003DHS, KE2008DHS, KE2014DHS) and 6 others.
+**Geographic** — `elderly` (range 500.0–12802.0).
 
-**Outcome / Measurement** — `value` (range 0.5–743.0), `istotal` (range 1.0–1.0).
+**Demographic** — `female_pop` (range 8224.0–309293.0), `female_u5` (range 1179.0–42669.0), `pop_u15` (range 6496.0–332085.0), `female_u15` (range 3224.0–156893.0).
 
-**Identifier / Metadata** — `dataid` (range 1361.0–824316.0), `indicatorid` (RH_DELP_C_DHF, CM_ECMR_C_IMR, CM_ECMR_C_U5M), `characteristicid` (range 1000.0–10000.0), `characteristiclabel` (Total, Total 15-49), `ispreferred` (range 0.0–1.0) and 3 others.
+**Identifier / Metadata** — `adm2_pcode` (KE027144, KE042242, KE032168), `adm_pcode` (KE027144, KE042242, KE032168), `esa_source` (HDX), `esa_processed` (2026-04-27).
 
-**Other** — `indicator` (Place of delivery: Health facility, Infant mortality rate, Under-five mortality rate), `precision` (range 0.0–1.0), `indicatororder` (range 11763080.0–260321010.0), `characteristicorder` (range 0.0–10000.0), `denominatorweighted` (range 549.0–37911.0) and 1 others.
+**Other** — `children_u5` (range 2398.0–85021.0).
 
 ---
 
@@ -129,33 +212,16 @@ train.head()
 
 | Column | Type | Null % | Range / Sample Values |
 |---|---|---|---|
-| `iso3` | object | 0.0% | KEN |
-| `dataid` | int64 | 0.0% | 1361.0 – 824316.0 (mean 439095.9476) |
-| `indicator` | object | 0.0% | Place of delivery: Health facility, Infant mortality rate, Under-five mortality rate |
-| `value` | float64 | 0.0% | 0.5 – 743.0 (mean 50.256) |
-| `precision` | int64 | 0.0% | 0.0 – 1.0 (mean 0.8325) |
-| `dhs_countrycode` | object | 0.0% | KE |
-| `countryname` | object | 0.0% | Kenya |
-| `surveyyear` | int64 | 0.0% | 1989.0 – 2022.0 (mean 2006.6963) |
-| `surveyid` | object | 0.0% | KE2003DHS, KE2008DHS, KE2014DHS |
-| `indicatorid` | object | 0.0% | RH_DELP_C_DHF, CM_ECMR_C_IMR, CM_ECMR_C_U5M |
-| `indicatororder` | int64 | 0.0% | 11763080.0 – 260321010.0 (mean 100874966.4921) |
-| `indicatortype` | object | 0.0% | I |
-| `characteristicid` | int64 | 0.0% | 1000.0 – 10000.0 (mean 2554.9738) |
-| `characteristicorder` | int64 | 0.0% | 0.0 – 10000.0 (mean 1727.7487) |
-| `characteristiccategory` | object | 0.0% | Total, Total 15-49 |
-| `characteristiclabel` | object | 0.0% | Total, Total 15-49 |
-| `byvariableid` | int64 | 0.0% | 0.0 – 631002.0 (mean 20550.1728) |
-| `byvariablelabel` | object | 68.6% | Five years preceding the survey, Ten years preceding the survey, Three years preceding the survey |
-| `istotal` | int64 | 0.0% | 1.0 – 1.0 (mean 1.0) |
-| `ispreferred` | int64 | 0.0% | 0.0 – 1.0 (mean 0.822) |
-| `sdrid` | object | 0.0% |  |
-| `surveyyearlabel` | object | 0.0% |  |
-| `surveytype` | object | 0.0% |  |
-| `denominatorweighted` | float64 | 31.9% | 549.0 – 37911.0 (mean 8269.8692) |
-| `denominatorunweighted` | float64 | 31.9% | 538.0 – 37911.0 (mean 8494.9077) |
-| `esa_source` | object | 0.0% |  |
-| `esa_processed` | object | 0.0% |  |
+| `adm2_pcode` | object | 0.0% | KE027144, KE042242, KE032168 |
+| `adm_pcode` | object | 0.0% | KE027144, KE042242, KE032168 |
+| `female_pop` | int64 | 0.0% | 8224.0 – 309293.0 (mean 94824.9517) |
+| `children_u5` | int64 | 0.0% | 2398.0 – 85021.0 (mean 26349.6517) |
+| `female_u5` | int64 | 0.0% | 1179.0 – 42669.0 (mean 13105.6621) |
+| `elderly` | int64 | 0.0% | 500.0 – 12802.0 (mean 5314.3034) |
+| `pop_u15` | int64 | 0.0% | 6496.0 – 332085.0 (mean 75028.2483) |
+| `female_u15` | int64 | 0.0% | 3224.0 – 156893.0 (mean 37273.269) |
+| `esa_source` | object | 0.0% | HDX |
+| `esa_processed` | object | 0.0% | 2026-04-27 |
 
 ---
 
@@ -163,33 +229,26 @@ train.head()
 
 | Column | Min | Max | Mean | Median |
 |---|---|---|---|---|
-| `dataid` | 1361.0 | 824316.0 | 439095.9476 | 462342.0 |
-| `value` | 0.5 | 743.0 | 50.256 | 39.2 |
-| `precision` | 0.0 | 1.0 | 0.8325 | 1.0 |
-| `surveyyear` | 1989.0 | 2022.0 | 2006.6963 | 2008.0 |
-| `indicatororder` | 11763080.0 | 260321010.0 | 100874966.4921 | 83566070.0 |
-| `characteristicid` | 1000.0 | 10000.0 | 2554.9738 | 1000.0 |
-| `characteristicorder` | 0.0 | 10000.0 | 1727.7487 | 0.0 |
-| `byvariableid` | 0.0 | 631002.0 | 20550.1728 | 0.0 |
-| `istotal` | 1.0 | 1.0 | 1.0 | 1.0 |
-| `ispreferred` | 0.0 | 1.0 | 0.822 | 1.0 |
-| `denominatorweighted` | 549.0 | 37911.0 | 8269.8692 | 5394.0 |
-| `denominatorunweighted` | 538.0 | 37911.0 | 8494.9077 | 5394.0 |
+| `female_pop` | 8224.0 | 309293.0 | 94824.9517 | 90518.5 |
+| `children_u5` | 2398.0 | 85021.0 | 26349.6517 | 24505.0 |
+| `female_u5` | 1179.0 | 42669.0 | 13105.6621 | 12246.5 |
+| `elderly` | 500.0 | 12802.0 | 5314.3034 | 5128.5 |
+| `pop_u15` | 6496.0 | 332085.0 | 75028.2483 | 67840.5 |
+| `female_u15` | 3224.0 | 156893.0 | 37273.269 | 34116.5 |
 
 ---
 
 ## Curation
 
-Raw data was downloaded from HDX via the CKAN API and converted to Parquet. Column names were lowercased and standardised to snake_case. Common missing-value markers (`N/A`, `null`, `none`, `-`, `unknown`, `no data`, `#N/A`) were unified to `NaN`. 4 column(s) with >80% missing values were removed: `regionid`, `cilow`, `cihigh`, `levelrank`. The dataset was split 80/20 into train and test partitions using a fixed random seed (42) and saved as Snappy-compressed Parquet.
+Raw data was downloaded from HDX via the CKAN API and converted to Parquet. Column names were lowercased and standardised to snake_case. Common missing-value markers (`N/A`, `null`, `none`, `-`, `unknown`, `no data`, `#N/A`) were unified to `NaN`. The dataset was split 80/20 into train and test partitions using a fixed random seed (42) and saved as Snappy-compressed Parquet.
 
 ---
 
 ## Limitations
 
-- Data originates from The DHS Program and has not been independently validated by ESA.
+- Data originates from HeiGIT (Heidelberg Institute for Geoinformation Technology) and has not been independently validated by ESA.
 - Automated cleaning cannot correct for misreported values, definitional inconsistencies, or sampling bias in the original collection.
-- The following columns have >20% missing values and should be treated with caution in modelling: `byvariablelabel`, `denominatorweighted`, `denominatorunweighted`.
-- Refer to the [original HDX dataset page](https://data.humdata.org/dataset/dhs-data-for-kenya) for the publisher's own methodology notes and caveats.
+- Refer to the [original HDX dataset page](https://data.humdata.org/dataset/kenya---risk-assessment-indicators) for the publisher's own methodology notes and caveats.
 
 ---
 
@@ -197,10 +256,10 @@ Raw data was downloaded from HDX via the CKAN API and converted to Parquet. Colu
 
 ```bibtex
 @dataset{hdx_africa_demographics_kenya,
-  title     = {Kenya - National Demographic and Health Data},
-  author    = {The DHS Program},
+  title     = {Kenya - Risk Assessment Indicators},
+  author    = {HeiGIT (Heidelberg Institute for Geoinformation Technology)},
   year      = {2026},
-  url       = {https://data.humdata.org/dataset/dhs-data-for-kenya},
+  url       = {https://data.humdata.org/dataset/kenya---risk-assessment-indicators},
   note      = {Repackaged for machine learning by Electric Sheep Africa (https://huggingface.co/electricsheepafrica)}
 }
 ```