Add README.md

README.md

@@ -1,45 +1,163 @@
 ---
+annotations_creators:
+- no-annotation
+language_creators:
+- found
+language:
+- en
+license: cc-by-sa-4.0
+multilinguality:
+- monolingual
+size_categories:
+- n<1K
+source_datasets:
+- original
+task_categories:
+- tabular-classification
+- tabular-regression
+- other
+task_ids: []
+tags:
+- africa
+- humanitarian
+- hdx
+- electric-sheep-africa
+- food-security
+- indicators
+- gin
+pretty_name: "Guinea - HungerMap data"
 dataset_info:
-  features:
-  - name: countrycode
-    dtype: string
-  - name: countryname
-    dtype: string
-  - name: adminlevel
-    dtype: string
-  - name: date
-    dtype: timestamp[ns]
-  - name: datatype
-    dtype: string
-  - name: fcs_people
-    dtype: int64
-  - name: fcs_prevalence
-    dtype: float64
-  - name: rcsi_people
-    dtype: int64
-  - name: rcsi_prevalence
-    dtype: float64
-  - name: health_access_people
-    dtype: int64
-  - name: health_access_prevalence
-    dtype: float64
-  - name: market_access_people
-    dtype: int64
-  - name: market_access_prevalence
-    dtype: float64
-  - name: esa_source
-    dtype: string
-  - name: esa_processed
-    dtype: string
   splits:
-  - name: train
-    num_bytes: 132
-    num_examples: 1
-  download_size: 6754
-  dataset_size: 132
-configs:
-- config_name: default
-  data_files:
-  - split: train
-    path: data/train-*
+    - name: train
+      num_examples: 0
+    - name: test
+      num_examples: 0
 ---
+
+# Guinea - HungerMap data
+
+**Publisher:** WFP - World Food Programme · **Source:** [HDX](https://data.humdata.org/dataset/wfp-hungermap-data-for-gin) · **License:** `cc-by-sa` · **Updated:** 2026-03-04
+
+---
+
+## Abstract
+
+HungerMapLIVE is the World Food Programme (WFP)’s global hunger monitoring system. It combines key metrics from various data sources – such as food security information, weather, population size, conflict, hazards, nutrition information and macro-economic data – to help assess, monitor and predict the magnitude and severity of hunger in near real-time. The resulting analysis is displayed on an interactive map that helps WFP staff, key decision makers and the broader humanitarian community to make more informed and timely decisions relating to food security.  
+  
+The platform covers 94 countries, including countries where WFP has operations as well as most lower and lower-middle income countries (as classified by the World Bank).
+
+Each row in this dataset represents country-level aggregates. Temporal coverage is indicated by the `date` column(s). Geographic scope: **GIN**.
+
+*Curated into ML-ready Parquet format by [Electric Sheep Africa](https://huggingface.co/electricsheepafrica).*
+
+---
+
+## Dataset Characteristics
+
+| | |
+|---|---|
+| **Domain** | Food security and nutrition |
+| **Unit of observation** | Country-level aggregates |
+| **Rows (total)** | 1 |
+| **Columns** | 15 (8 numeric, 6 categorical, 1 datetime) |
+| **Train split** | 0 rows |
+| **Test split** | 0 rows |
+| **Geographic scope** | GIN |
+| **Publisher** | WFP - World Food Programme |
+| **HDX last updated** | 2026-03-04 |
+
+---
+
+## Variables
+
+**Geographic** — `countrycode` (GIN), `countryname` (Guinea), `adminlevel` (national), `datatype` (SURVEY).
+
+**Temporal** — `date`.
+
+**Identifier / Metadata** — `esa_source` (HDX), `esa_processed` (2026-04-06).
+
+**Other** — `fcs_people` (range 6992096.0–6992096.0), `fcs_prevalence` (range 0.5632–0.5632), `rcsi_people` (range 4079611.0–4079611.0), `rcsi_prevalence` (range 0.3286–0.3286), `health_access_people` (range 5746383.0–5746383.0) and 3 others.
+
+---
+
+## Quick Start
+
+```python
+from datasets import load_dataset
+
+ds    = load_dataset("electricsheepafrica/africa-wfp-hungermap-data-for-gin")
+train = ds["train"].to_pandas()
+test  = ds["test"].to_pandas()
+
+print(train.shape)
+train.head()
+```
+
+---
+
+## Schema
+
+| Column | Type | Null % | Range / Sample Values |
+|---|---|---|---|
+| `countrycode` | object | 0.0% | GIN |
+| `countryname` | object | 0.0% | Guinea |
+| `adminlevel` | object | 0.0% | national |
+| `date` | datetime64[ns] | 0.0% |  |
+| `datatype` | object | 0.0% | SURVEY |
+| `fcs_people` | int64 | 0.0% | 6992096.0 – 6992096.0 (mean 6992096.0) |
+| `fcs_prevalence` | float64 | 0.0% | 0.5632 – 0.5632 (mean 0.5632) |
+| `rcsi_people` | int64 | 0.0% | 4079611.0 – 4079611.0 (mean 4079611.0) |
+| `rcsi_prevalence` | float64 | 0.0% | 0.3286 – 0.3286 (mean 0.3286) |
+| `health_access_people` | int64 | 0.0% | 5746383.0 – 5746383.0 (mean 5746383.0) |
+| `health_access_prevalence` | float64 | 0.0% | 0.8465 – 0.8465 (mean 0.8465) |
+| `market_access_people` | int64 | 0.0% | 6067543.0 – 6067543.0 (mean 6067543.0) |
+| `market_access_prevalence` | float64 | 0.0% | 0.4888 – 0.4888 (mean 0.4888) |
+| `esa_source` | object | 0.0% | HDX |
+| `esa_processed` | object | 0.0% | 2026-04-06 |
+
+---
+
+## Numeric Summary
+
+| Column | Min | Max | Mean | Median |
+|---|---|---|---|---|
+| `fcs_people` | 6992096.0 | 6992096.0 | 6992096.0 | 6992096.0 |
+| `fcs_prevalence` | 0.5632 | 0.5632 | 0.5632 | 0.5632 |
+| `rcsi_people` | 4079611.0 | 4079611.0 | 4079611.0 | 4079611.0 |
+| `rcsi_prevalence` | 0.3286 | 0.3286 | 0.3286 | 0.3286 |
+| `health_access_people` | 5746383.0 | 5746383.0 | 5746383.0 | 5746383.0 |
+| `health_access_prevalence` | 0.8465 | 0.8465 | 0.8465 | 0.8465 |
+| `market_access_people` | 6067543.0 | 6067543.0 | 6067543.0 | 6067543.0 |
+| `market_access_prevalence` | 0.4888 | 0.4888 | 0.4888 | 0.4888 |
+
+---
+
+## 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`. 2 column(s) with >80% missing values were removed: `adminone`, `population`. 1 column(s) were cast from string to numeric or datetime based on parse-success rate (>85% threshold). 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 WFP - World Food Programme and has not been independently validated by ESA.
+- Automated cleaning cannot correct for misreported values, definitional inconsistencies, or sampling bias in the original collection.
+- Refer to the [original HDX dataset page](https://data.humdata.org/dataset/wfp-hungermap-data-for-gin) for the publisher's own methodology notes and caveats.
+
+---
+
+## Citation
+
+```bibtex
+@dataset{hdx_africa_wfp_hungermap_data_for_gin,
+  title     = {Guinea - HungerMap data},
+  author    = {WFP - World Food Programme},
+  year      = {2026},
+  url       = {https://data.humdata.org/dataset/wfp-hungermap-data-for-gin},
+  note      = {Repackaged for machine learning by Electric Sheep Africa (https://huggingface.co/electricsheepafrica)}
+}
+```
+
+---
+
+*[Electric Sheep Africa](https://huggingface.co/electricsheepafrica) — Africa's ML dataset infrastructure. Lagos, Nigeria.*