electricsheepafrica/africa-wfp-food-prices-for-madagascar

Madagascar - Food Prices

Publisher: WFP - World Food Programme · Source: HDX · License: cc-by-igo · Updated: 2026-04-05

---

Abstract

This dataset contains Food Prices data for Madagascar, sourced from the World Food Programme Price Database. The World Food Programme Price Database covers foods such as maize, rice, beans, fish, and sugar for 98 countries and some 3000 markets. It is updated weekly but contains to a large extent monthly data. The data goes back as far as 1992 for a few countries, although many countries started reporting from 2003 or thereafter.

Each row in this dataset represents subnational administrative unit observations. Temporal coverage is indicated by the date column(s). Geographic scope: MDG.

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

---

Dataset Characteristics

Domain	Food security and nutrition
Unit of observation	Subnational administrative unit observations
Rows (total)	23,114
Columns	18 (6 numeric, 11 categorical, 1 datetime)
Train split	18,491 rows
Test split	4,622 rows
Geographic scope	MDG
Publisher	WFP - World Food Programme
HDX last updated	2026-04-05

---

Variables

Geographic — admin1 (Atsimo Andrefana, Androy, Anosy), admin2 (Betioky Atsimo, Ampanihy Ouest, Ambovombe-Androy), latitude (range -25.54–-12.28), longitude (range 43.31–50.17), category (cereals and tubers, oil and fats, pulses and nuts) and 4 others.

Temporal — date.

Outcome / Measurement — priceflag (aggregate, actual), price (range 220.0–16000.0), usdprice (range 0.1–3.73).

Identifier / Metadata — market_id (range 732.0–11286.0), esa_source (HDX), esa_processed.

Other — market (Amboasary Sud, Tsihombe, Ampanihy), unit (KG, L).

---

Quick Start

from datasets import load_dataset

ds    = load_dataset("electricsheepafrica/africa-wfp-food-prices-for-madagascar")
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%
admin1	object	0.0%	Atsimo Andrefana, Androy, Anosy
admin2	object	0.0%	Betioky Atsimo, Ampanihy Ouest, Ambovombe-Androy
market	object	0.0%	Amboasary Sud, Tsihombe, Ampanihy
market_id	int64	0.0%	732.0 – 11286.0 (mean 5547.3975)
latitude	float64	0.0%	-25.54 – -12.28 (mean -22.6118)
longitude	float64	0.0%	43.31 – 50.17 (mean 46.0967)
category	object	0.0%	cereals and tubers, oil and fats, pulses and nuts
commodity	object	0.0%	Rice (local), Rice (imported), Maize (crushed)
commodity_id	int64	0.0%	51.0 – 1332.0 (mean 345.2263)
unit	object	0.0%	KG, L
priceflag	object	0.0%	aggregate, actual
pricetype	object	0.0%	Retail, Producer
currency	object	0.0%	MGA
price	float64	0.0%	220.0 – 16000.0 (mean 3690.1987)
usdprice	float64	0.0%	0.1 – 3.73 (mean 0.9432)
esa_source	object	0.0%	HDX
esa_processed	object	0.0%

---

Numeric Summary

Column	Min	Max	Mean	Median
market_id	732.0	11286.0	5547.3975	6568.0
latitude	-25.54	-12.28	-22.6118	-23.58
longitude	43.31	50.17	46.0967	45.86
commodity_id	51.0	1332.0	345.2263	283.0
price	220.0	16000.0	3690.1987	2800.0
usdprice	0.1	3.73	0.9432	0.72

---

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) 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 for the publisher's own methodology notes and caveats.

---

Citation

@dataset{hdx_africa_wfp_food_prices_for_madagascar,
  title     = {Madagascar - Food Prices},
  author    = {WFP - World Food Programme},
  year      = {2026},
  url       = {https://data.humdata.org/dataset/wfp-food-prices-for-madagascar},
  note      = {Repackaged for machine learning by Electric Sheep Africa (https://huggingface.co/electricsheepafrica)}
}

---

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