electricsheepafrica/africa-sahel-prediction-dead-animals-by-ach-gis4tech

Sahel Prediction Death Animals

Publisher: Acción contra el hambre - GIS4tech · Source: HDX · License: cc-by · Updated: 2026-04-01

---

Abstract

The model trains with the actual data of the variables to be predicted and uses their correlation with agroclimatic indicators, biomass, and violence to predict said variable where there is no actual data, at the level of Commune/Municipality for all of Senegal and Mauritania.

This information belongs to the Food and Nutrition Security Monitoring and Prediction System in the Sahel (PREDISAN) Project, funded by the Agencia Andaluza de Cooperación Internacional para el Desarrollo (AACID) and the University of Granada (UGR). The PREDISAN AI-SAHEL Project focuses on the Monitoring and Prediction System for Humanitarian Vulnerability of Pastoral and Agro-pastoral Populations in the Western Sahel, based on GIS Analysis and Artificial Intelligence.

For more information, contact GIS4Tech at info@gis4tech.com or visit the PREDISAN platform at https://predisan.gis4tech.com/sahel.

Each row in this dataset represents country-level aggregates. Temporal coverage is indicated by the date column(s). Geographic scope: MRT, SEN.

Curated into ML-ready Parquet format by Electric Sheep Africa.

---

Dataset Characteristics

Domain	Food security and nutrition
Unit of observation	Country-level aggregates
Rows (total)	36,059
Columns	9 (3 numeric, 5 categorical, 1 datetime)
Train split	28,847 rows
Test split	7,211 rows
Geographic scope	MRT, SEN
Publisher	Acción contra el hambre - GIS4tech
HDX last updated	2026-04-01

---

Variables

Geographic — country (Mauritania, Senegal), municipality (Mabrouk, Lexeiba, Nouamghar).

Temporal — date.

Identifier / Metadata — esa_source (HDX), esa_processed (2026-04-12).

Other — department (Sélibaby, Néma, M'Bout), prob_dead_animals_0 (range 0.1159–0.999), prob_dead_animals_1 (range 0.001–0.8841), dead_animals (range 0.0–1.0).

---

Quick Start

from datasets import load_dataset

ds    = load_dataset("electricsheepafrica/africa-sahel-prediction-dead-animals-by-ach-gis4tech")
train = ds["train"].to_pandas()
test  = ds["test"].to_pandas()

print(train.shape)
train.head()

---

Schema

Column	Type	Null %	Range / Sample Values
date	datetime64[ns]	0.0%
country	object	0.0%	Mauritania, Senegal
department	object	0.0%	Sélibaby, Néma, M'Bout
municipality	object	0.0%	Mabrouk, Lexeiba, Nouamghar
prob_dead_animals_0	float64	0.0%	0.1159 – 0.999 (mean 0.8897)
prob_dead_animals_1	float64	0.0%	0.001 – 0.8841 (mean 0.1103)
dead_animals	int64	0.0%	0.0 – 1.0 (mean 0.0375)
esa_source	object	0.0%	HDX
esa_processed	object	0.0%	2026-04-12

---

Numeric Summary

Column	Min	Max	Mean	Median
prob_dead_animals_0	0.1159	0.999	0.8897	0.9423
prob_dead_animals_1	0.001	0.8841	0.1103	0.0577
dead_animals	0.0	1.0	0.0375	0.0

---

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. 1,830 exact duplicate rows were removed. 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 Acción contra el hambre - GIS4tech and has not been independently validated by ESA.
• Automated cleaning cannot correct for misreported values, definitional inconsistencies, or sampling bias in the original collection.
• This dataset spans 2 countries; geographic and methodological inconsistencies across national boundaries may affect cross-country comparability.
• Refer to the original HDX dataset page for the publisher's own methodology notes and caveats.

---

Citation

@dataset{hdx_africa_sahel_prediction_dead_animals_by_ach_gis4tech,
  title     = {Sahel Prediction Death Animals},
  author    = {Acción contra el hambre - GIS4tech},
  year      = {2026},
  url       = {https://data.humdata.org/dataset/sahel-prediction-dead-animals-by-ach-gis4tech},
  note      = {Repackaged for machine learning by Electric Sheep Africa (https://huggingface.co/electricsheepafrica)}
}

---

Electric Sheep Africa — Africa's ML dataset infrastructure. Lagos, Nigeria.