electricsheepafrica/africa-covid-19-impacts-africa

---
annotations_creators:
- no-annotation
language_creators:
- found
language:
- en
license: cc-by-4.0
multilinguality:
- monolingual
size_categories:
- 1K<n<10K
source_datasets:
- original
task_categories:
- tabular-classification
- other
task_ids: []
tags:
- africa
- humanitarian
- hdx
- electric-sheep-africa
- covid-19
- epidemics-outbreaks
- health
- ben
- civ
- cod
- gha
- ken
pretty_name: "Perceptions and impact of Coronavirus in Sub-Saharan African countries"
dataset_info:
splits:
- name: train
num_examples: 3189
- name: test
num_examples: 797
---

# Perceptions and impact of Coronavirus in Sub-Saharan African countries

**Publisher:** Mobile Accord, Inc. (GeoPoll) · **Source:** [HDX](https://data.humdata.org/dataset/covid-19-impacts-africa) · **License:** `cc-by` · **Updated:** 2025-09-26

---

## Abstract

This data and report examine perceptions and the impact of COVID-19 in 12 countries throughout sub-Saharan Africa. Topics covered include greatest concerns surrounding coronavirus, preventative measures being taken, changes in food market operability and food security, consumer behavior changes, and trust in governments to prevent the spread of coronavirus. This dataset includes data from 10 of the markets. Please contact us for access to data from all markets, the questionnaire, and with any other questions.

Each row in this dataset represents first-level administrative unit observations. Temporal coverage is indicated by the `survey_date` column(s). Geographic scope: **BEN, CIV, COD, GHA, KEN, MOZ, NGA, RWA, and 4 others**.

*Curated into ML-ready Parquet format by [Electric Sheep Africa](https://huggingface.co/electricsheepafrica).*

---

## Dataset Characteristics

| | |
|---|---|
| **Domain** | Food security and nutrition |
| **Unit of observation** | First-level administrative unit observations |
| **Rows (total)** | 3,987 |
| **Columns** | 64 (9 numeric, 28 categorical, 1 datetime) |
| **Train split** | 3,189 rows |
| **Test split** | 797 rows |
| **Geographic scope** | BEN, CIV, COD, GHA, KEN, MOZ, NGA, RWA, and 4 others |
| **Publisher** | Mobile Accord, Inc. (GeoPoll) |
| **HDX last updated** | 2025-09-26 |

---

## Variables

**Geographic** — `survey_date`, `country` (Kenya, Nigeria, Ghana), `admin1` (Lusaka, Abidjan, Kigali), `birthyear` (range 1928.0–2004.0), `preventativemeasures_avoiding_public_places` and 7 others.

**Demographic** — `gender_weights` (range 0.7608–1.4586), `age_group_weights` (range 0.619–1.8537), `gender` (Male, Female), `age_group` (15-25, 26-35, 36+), `age` (range 15.0–91.0) and 1 others.

**Outcome / Measurement** — `total_weights` (range 0.1396–6.4719), `foodamount`.

**Identifier / Metadata** — `preventativemeasures_avoiding_public_transport`, `informationsources_newspapers`, `informationsources_other`, `informationsources_radio`, `informationsources_social_media` and 3 others.

**Other** — `adm1_weights` (range 0.2219–3.7881), `sec` (C1, B, A), `awareness` (Yes), `urban_rural` (Urban area, Rural area), `biggestchallenge` (Money, Corona virus, Coronavirus) and 31 others.

---

## Quick Start

```python
from datasets import load_dataset

ds = load_dataset("electricsheepafrica/africa-covid-19-impacts-africa")
train = ds["train"].to_pandas()
test = ds["test"].to_pandas()

print(train.shape)
train.head()
```

---

## Schema

| Column | Type | Null % | Range / Sample Values |
|---|---|---|---|
| `gender_weights` | float64 | 0.0% | 0.7608 – 1.4586 (mean 1.0) |
| `age_group_weights` | float64 | 0.0% | 0.619 – 1.8537 (mean 1.0001) |
| `adm1_weights` | float64 | 0.0% | 0.2219 – 3.7881 (mean 0.9987) |
| `total_weights` | float64 | 0.0% | 0.1396 – 6.4719 (mean 0.9881) |
| `survey_date` | datetime64[ns] | 0.0% | |
| `country` | object | 0.0% | Kenya, Nigeria, Ghana |
| `gender` | object | 0.0% | Male, Female |
| `age_group` | object | 0.0% | 15-25, 26-35, 36+ |
| `admin1` | object | 0.0% | Lusaka, Abidjan, Kigali |
| `sec` | object | 0.0% | C1, B, A |
| `age` | int64 | 0.0% | 15.0 – 91.0 (mean 31.2904) |
| `birthyear` | int64 | 0.0% | 1928.0 – 2004.0 (mean 1987.7096) |
| `awareness` | object | 0.0% | Yes |
| `urban_rural` | object | 0.0% | Urban area, Rural area |
| `biggestchallenge` | object | 49.9% | Money, Corona virus, Coronavirus |
| `levelconcern` | int64 | 0.0% | 1.0 – 5.0 (mean 4.3128) |
| `concerns` | object | 0.0% | Contracting the disease, Economic impact, Global infections |
| `riskawareness` | object | 0.0% | Yes, No |
| `tested` | object | 0.0% | |
| `virusprevention` | object | 0.0% | |
| `preventativemeasures_avoiding_public_places` | bool | 0.0% | |
| `preventativemeasures_avoiding_public_transport` | bool | 0.0% | |
| `preventativemeasures_increasing_hygiene` | bool | 0.0% | |
| `preventativemeasures_other` | bool | 0.0% | |
| `preventativemeasures_working_from_home` | bool | 0.0% | |
| `handwashing` | object | 0.0% | |
| `socialdistancing` | object | 0.0% | |
| `healthbehavior` | object | 0.0% | |
| `economicimpact` | int64 | 0.0% | 1.0 – 5.0 (mean 4.2854) |
| `marketoperability` | object | 0.0% | |
| `foodlocations` | object | 0.0% | |
| `foodlocations2` | object | 48.3% | |
| `foodshopping` | object | 0.0% | |
| `foodamount` | object | 0.0% | |
| `foodworry` | object | 0.0% | |
| `brandpurchase` | object | 20.3% | |
| `nonessentialitems` | object | 0.0% | |
| `governmenttrust` | int64 | 0.0% | 1.0 – 5.0 (mean 3.4542) |
| `commercialtrust_banks` | bool | 0.0% | |
| `commercialtrust_brands` | bool | 0.0% | |
| `commercialtrust_other` | bool | 0.0% | |
| `commercialtrust_retailers` | bool | 0.0% | |
| `commercialtrust_telecommuncations` | bool | 0.0% | |
| `initiativetaken_limited_store_traffic` | bool | 0.0% | |
| `initiativetaken_none` | bool | 0.0% | |
| `initiativetaken_sectioned_intercepts` | bool | 0.0% | |
| `initiativetaken_staff_in_protective_gear` | bool | 0.0% | |
| `groups_brands` | bool | 0.0% | |
| `groups_governments` | bool | 0.0% | |
| `groups_ngos` | bool | 0.0% | |
| `groups_private_sector` | bool | 0.0% | |
| `groups_retailers` | bool | 0.0% | |
| `informationsources_friends_family` | bool | 0.0% | |
| `informationsources_government_messages` | bool | 0.0% | |
| `informationsources_newspapers` | bool | 0.0% | |
| `informationsources_other` | bool | 0.0% | |
| `informationsources_radio` | bool | 0.0% | |
| `informationsources_social_media` | bool | 0.0% | |
| `informationsources_tv` | bool | 0.0% | |
| `socialmedia` | object | 59.3% | |
| `mediaconsumption` | object | 0.0% | |
| `staypositive` | object | 49.9% | |
| `esa_source` | object | 0.0% | |
| `esa_processed` | object | 0.0% | |

---

## Numeric Summary

| Column | Min | Max | Mean | Median |
|---|---|---|---|---|
| `gender_weights` | 0.7608 | 1.4586 | 1.0 | 1.0 |
| `age_group_weights` | 0.619 | 1.8537 | 1.0001 | 1.0 |
| `adm1_weights` | 0.2219 | 3.7881 | 0.9987 | 1.0 |
| `total_weights` | 0.1396 | 6.4719 | 0.9881 | 1.0 |
| `age` | 15.0 | 91.0 | 31.2904 | 29.0 |
| `birthyear` | 1928.0 | 2004.0 | 1987.7096 | 1990.0 |
| `levelconcern` | 1.0 | 5.0 | 4.3128 | 5.0 |
| `economicimpact` | 1.0 | 5.0 | 4.2854 | 5.0 |
| `governmenttrust` | 1.0 | 5.0 | 3.4542 | 4.0 |

---

## 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 exact duplicate rows were removed. 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 Mobile Accord, Inc. (GeoPoll) 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: `biggestchallenge`, `foodlocations2`, `brandpurchase`, `socialmedia`, `staypositive`.
- This dataset spans 12 countries; geographic and methodological inconsistencies across national boundaries may affect cross-country comparability.
- Refer to the [original HDX dataset page](https://data.humdata.org/dataset/covid-19-impacts-africa) for the publisher's own methodology notes and caveats.

---

## Citation

```bibtex
@dataset{hdx_africa_covid_19_impacts_africa,
title = {Perceptions and impact of Coronavirus in Sub-Saharan African countries},
author = {Mobile Accord, Inc. (GeoPoll)},
year = {2025},
url = {https://data.humdata.org/dataset/covid-19-impacts-africa},
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.*