electricsheepafrica/africa-wash-ethiopia

Water Point Data Exchange (ETH)

Publisher: Water Point Data Exchange · Source: HDX · License: Creative Commons Attribution Share-Alike · Updated: 2025-12-14

---

Abstract

WPdx is a platform to compile crowdsourced data focused on rural water points (wells, springs, tapstands) with contributions from governments, NGOs, and researchers. Shared data is cleaned and harmonized using the WPdx Data Standard to create a robust analysis-ready dataset. There are two primary datasets available from WPdx. The first is WPdx-Basic, which includes all records shared with the platform. The second, WPdx-Plus is a subset of the Basic dataset which focuses on countries where district and/or national data is available and undergoes additional cleaning. WPdx-Plus is used as an input to the WPdx Decision Support tools and is the basis for the datasets posted here. The decision support tools provide insights on rural basic water service access and recommendations regarding prioritized water point repair, highlight areas with apparent service gaps, and identify water points which are at high risk of failure. For each country where WPdx data is available, there are two datasets available for download are described below.

• The wpdx_adm_region_analysis file provides an overview of the population served, unserved and uncharted for each available administrative level. The file also includes a data quality analysis based on water point record age.
• The wpdx_water_points file provides water point level details for each point shared with WPdx including all the WPdx data standard parameters plus results from the WPdx Rehab Priority, Service Gap/New Construction and Status Prediction analyses.
• A data dictionary describing each parameter included in each file is available in the Data and Resources section.
  To share data with the WPdx platform, please visit (www.waterpointdata.org/share-data) and/or reach out to info@waterpointdata.org

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

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

---

Dataset Characteristics

Domain	Demographics and population
Unit of observation	Country-level aggregates
Rows (total)	24,489
Columns	40 (15 numeric, 23 categorical, 2 datetime)
Train split	19,591 rows
Test split	4,897 rows
Geographic scope	ETH
Publisher	Water Point Data Exchange
HDX last updated	2025-12-14

---

Variables

Geographic — lat_deg (range 3.4562–14.6071), lon_deg (range 34.4713–47.3185), clean_country_id (ETH, #country+code+v_iso3), clean_country_name (Ethiopia, #country+name), activity_id (Kef Mender, 182535349, 181732293) and 13 others.

Temporal — report_date, created_timestamp.

Demographic — management_clean, usage_cap (range 50.0–300.0).

Identifier / Metadata — status_id (Yes, No, Unknown), source (COWASH, Amhara Region, Farta Woreda Water Office), water_source_clean (Protected Well, Borehole/Tubewell, Protected Spring), scheme_id, dataset_title and 2 others.

Other — water_tech_clean (Hand Pump, Hand Pump - Afridev, Hand Pump - Rope), clean_adm1 (Amhara, Oromia, Somali), clean_adm2 (South Gondar, West Gojam, Central Gondar), clean_adm3 (Dera (AM), Farta, Debub Mecha), installer and 6 others.

---

Quick Start

from datasets import load_dataset

ds    = load_dataset("electricsheepafrica/africa-wash-ethiopia")
train = ds["train"].to_pandas()
test  = ds["test"].to_pandas()

print(train.shape)
train.head()

---

Schema

Column	Type	Null %	Range / Sample Values
lat_deg	float64	0.0%	3.4562 – 14.6071 (mean 10.822)
lon_deg	float64	0.0%	34.4713 – 47.3185 (mean 38.0407)
status_id	object	0.0%	Yes, No, Unknown
report_date	datetime64[ns]	0.0%
source	object	0.0%	COWASH, Amhara Region, Farta Woreda Water Office
water_source_clean	object	0.7%	Protected Well, Borehole/Tubewell, Protected Spring
water_tech_clean	object	25.5%	Hand Pump, Hand Pump - Afridev, Hand Pump - Rope
clean_country_id	object	0.0%	ETH, #country+code+v_iso3
clean_country_name	object	0.0%	Ethiopia, #country+name
clean_adm1	object	0.0%	Amhara, Oromia, Somali
clean_adm2	object	0.0%	South Gondar, West Gojam, Central Gondar
clean_adm3	object	0.0%	Dera (AM), Farta, Debub Mecha
activity_id	object	77.9%	Kef Mender, 182535349, 181732293
scheme_id	object	78.0%
wpdx_id	object	0.0%
install_year	float64	20.6%	1902.0 – 2027.0 (mean 2007.878)
installer	object	77.7%
management_clean	object	25.1%
status_clean	object	0.0%
notes	object	25.7%
local_population	float64	1.0%	0.0 – 133575.0 (mean 1506.7955)
assigned_population	float64	1.0%	0.0 – 133575.0 (mean 543.459)
facility_type	object	0.0%
water_source_category	object	0.7%
water_tech_category	object	27.5%
created_timestamp	datetime64[ns]	0.0%
dataset_title	object	0.0%
usage_cap	float64	0.8%	50.0 – 300.0 (mean 298.0544)
criticality	float64	2.6%	0.0065 – 1.0 (mean 0.3944)
pressure	float64	2.6%	0.0033 – 445.25 (mean 1.8698)
distance_to_primary	float64	0.0%	2.953 – 202605.8371 (mean 56673.5096)
distance_to_secondary	float64	0.0%	0.6175 – 184552.1669 (mean 26632.2669)
distance_to_tertiary	float64	0.0%	0.0147 – 130490.4058 (mean 8684.356)
distance_to_city	float64	0.0%	548.7957 – 320039.7785 (mean 65596.7552)
distance_to_town	float64	0.0%	14.4077 – 163129.7517 (mean 18392.6123)
is_urban	object	0.0%
days_since_report	float64	0.0%	1096.0 – 9854.0 (mean 2176.4189)
staleness	float64	0.0%	4.4775 – 70.7885 (mean 52.1747)
esa_source	object	0.0%
esa_processed	object	0.0%

---

Numeric Summary

Column	Min	Max	Mean	Median
lat_deg	3.4562	14.6071	10.822	11.34
lon_deg	34.4713	47.3185	38.0407	37.8263
install_year	1902.0	2027.0	2007.878	2008.0
local_population	0.0	133575.0	1506.7955	964.0
assigned_population	0.0	133575.0	543.459	242.0
usage_cap	50.0	300.0	298.0544	300.0
criticality	0.0065	1.0	0.3944	0.2959
pressure	0.0033	445.25	1.8698	0.8367
distance_to_primary	2.953	202605.8371	56673.5096	55860.18
distance_to_secondary	0.6175	184552.1669	26632.2669	18764.1813
distance_to_tertiary	0.0147	130490.4058	8684.356	5868.7505
distance_to_city	548.7957	320039.7785	65596.7552	63057.9474
distance_to_town	14.4077	163129.7517	18392.6123	16202.7047
days_since_report	1096.0	9854.0	2176.4189	1801.0
staleness	4.4775	70.7885	52.1747	56.6828

---

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. 16 column(s) with >80% missing values were removed: clean_adm4, rehab_year, rehabilitator, pay_clean, fecal_coliform_presence, fecal_coliform_value.... 17 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 Water Point Data Exchange 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: water_tech_clean, activity_id, scheme_id, install_year, installer, management_clean, notes, water_tech_category.
• Refer to the original HDX dataset page for the publisher's own methodology notes and caveats.

---

Citation

@dataset{hdx_africa_wash_ethiopia,
  title     = {Water Point Data Exchange (ETH)},
  author    = {Water Point Data Exchange},
  year      = {2025},
  url       = {https://data.humdata.org/dataset/wpdx_eth},
  note      = {Repackaged for machine learning by Electric Sheep Africa (https://huggingface.co/electricsheepafrica)}
}

---

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