electricsheepafrica/africa-darfur-damaged-and-destroyed-villages

Darfur Damaged and Destroyed Villages

Publisher: U.S. Department of State - Humanitarian Information Unit (inactive) · Source: HDX · License: other-pd-nr · Updated: 2025-05-05

---

Abstract

The Darfur Damaged and Destroyed Villages dataset describes the condition of villages in the Darfur region of Sudan that the U.S. Government has confirmed as either damaged or destroyed between the time period February 2003 to December 2010. Additionally, villages that are confirmed to have No Damage are also reported.

Each row in this dataset represents geolocated point observations. Data was last updated on HDX on 2025-05-05. Geographic scope: SDN.

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

---

Dataset Characteristics

Domain	Conflict and security
Unit of observation	Geolocated point observations
Rows (total)	9,183
Columns	14 (9 numeric, 5 categorical, 0 datetime)
Train split	7,346 rows
Test split	1,836 rows
Geographic scope	SDN
Publisher	U.S. Department of State - Humanitarian Information Unit (inactive)
HDX last updated	2025-05-05

---

Variables

Geographic — latitude (range 91624.0–14243299.0), lat_dd (range 9.2733–15.5889), longitude (range 215006.0–25534651.0), long_dd (range 21.835–27.39), yr_confirm (range 2003.0–2010.0) and 4 others.

Identifier / Metadata — name (HASHABA, KUMA, HILLET IBRAHIM), esa_source (HDX), esa_processed (2026-04-05).

Other — status (NO DAMAGE, DESTROYED, DAMAGED), structures ( , ~25 of ~25, ~50 of ~50).

---

Quick Start

from datasets import load_dataset

ds    = load_dataset("electricsheepafrica/africa-darfur-damaged-and-destroyed-villages")
train = ds["train"].to_pandas()
test  = ds["test"].to_pandas()

print(train.shape)
train.head()

---

Schema

Column	Type	Null %	Range / Sample Values
name	object	68.8%	HASHABA, KUMA, HILLET IBRAHIM
latitude	float64	1.5%	91624.0 – 14243299.0 (mean 150722.5973)
lat_dd	float64	0.0%	9.2733 – 15.5889 (mean 12.8066)
longitude	float64	1.5%	215006.0 – 25534651.0 (mean 289943.1203)
long_dd	float64	0.0%	21.835 – 27.39 (mean 24.4526)
status	object	0.0%	NO DAMAGE, DESTROYED, DAMAGED
structures	object	58.7%	, ~25 of ~25, ~50 of ~50
yr_confirm	float64	71.8%	2003.0 – 2010.0 (mean 2004.7153)
aprox_str1	float64	78.1%	5.0 – 1500.0 (mean 105.8607)
aprox_str2	float64	70.5%	10.0 – 4000.0 (mean 117.8581)
yr_range1	float64	66.3%	2003.0 – 2010.0 (mean 2004.6276)
yr_range2	float64	66.3%	2003.0 – 2010.0 (mean 2004.7904)
esa_source	object	0.0%	HDX
esa_processed	object	0.0%	2026-04-05

---

Numeric Summary

Column	Min	Max	Mean	Median
latitude	91624.0	14243299.0	150722.5973	125656.0
lat_dd	9.2733	15.5889	12.8066	12.9725
longitude	215006.0	25534651.0	289943.1203	243237.0
long_dd	21.835	27.39	24.4526	24.5558
yr_confirm	2003.0	2010.0	2004.7153	2004.0
aprox_str1	5.0	1500.0	105.8607	50.0
aprox_str2	10.0	4000.0	117.8581	75.0
yr_range1	2003.0	2010.0	2004.6276	2004.0
yr_range2	2003.0	2010.0	2004.7904	2004.0

---

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 column(s) with >80% missing values were removed: sec_town. 3 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 U.S. Department of State - Humanitarian Information Unit (inactive) 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: name, structures, yr_confirm, aprox_str1, aprox_str2, yr_range1, yr_range2.
• Refer to the original HDX dataset page for the publisher's own methodology notes and caveats.

---

Citation

@dataset{hdx_africa_darfur_damaged_and_destroyed_villages,
  title     = {Darfur Damaged and Destroyed Villages},
  author    = {U.S. Department of State - Humanitarian Information Unit (inactive)},
  year      = {2025},
  url       = {https://data.humdata.org/dataset/darfur-damaged-and-destroyed-villages},
  note      = {Repackaged for machine learning by Electric Sheep Africa (https://huggingface.co/electricsheepafrica)}
}

---

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