electricsheepafrica/africa-ucdp-data-for-south-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
- conflict-violence
- hxl
- zaf
pretty_name: "South Africa - Data on Conflict Events"
dataset_info:
splits:
- name: train
num_examples: 2316
- name: test
num_examples: 579
---

# South Africa - Data on Conflict Events

**Publisher:** HDX · **Source:** [HDX](https://data.humdata.org/dataset/ucdp-data-for-south-africa) · **License:** `cc-by-igo` · **Updated:** 2026-04-03

---

## Abstract

This dataset is UCDP's most disaggregated dataset, covering individual events of organized violence (phenomena of lethal violence occurring at a given time and place). These events are sufficiently fine-grained to be geo-coded down to the level of individual villages, with temporal durations disaggregated to single, individual days.
Sundberg, Ralph, and Erik Melander, 2013, “Introducing the UCDP Georeferenced Event Dataset”, Journal of Peace Research, vol.50, no.4, 523-532
Högbladh Stina, 2019, “UCDP GED Codebook version 19.1”, Department of Peace and Conflict Research, Uppsala University

Each row in this dataset represents first-level administrative unit observations. Temporal coverage is indicated by the `date_start`, `date_end` column(s). Geographic scope: **ZAF**.

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

---

## Dataset Characteristics

| | |
|---|---|
| **Domain** | Conflict and security |
| **Unit of observation** | First-level administrative unit observations |
| **Rows (total)** | 2,895 |
| **Columns** | 47 (26 numeric, 18 categorical, 2 datetime) |
| **Train split** | 2,316 rows |
| **Test split** | 579 rows |
| **Geographic scope** | ZAF |
| **Publisher** | HDX |
| **HDX last updated** | 2026-04-03 |

---

## Variables

**Geographic** — `year` (range 1989.0–2024.0), `active_year`, `type_of_violence` (range 1.0–3.0), `dyad_dset_id` (range 102.0–16983.0), `dyad_new_id` (range 641.0–16983.0) and 9 others.

**Temporal** — `date_prec` (range 1.0–5.0), `date_start`, `date_end`.

**Outcome / Measurement** — `number_of_sources` (range -1.0–6.0), `deaths_a` (range 0.0–33.0), `deaths_b`, `deaths_civilians`, `deaths_unknown`.

**Identifier / Metadata** — `id` (range 13263.0–551051.0), `relid` (SAF-1989-1-516-8, SAF-1992-2-312-8, SAF-1992-2-312-300), `code_status` (Clear), `conflict_dset_id` (range 102.0–16983.0), `conflict_new_id` (range 298.0–15477.0) and 12 others.

**Other** — `where_prec` (range 1.0–6.0), `where_description` (Mpumalanga town (township), Soweto town (township), Alexandra town (township) (outskirts of Johannesburg town)), `adm_1`, `adm_2`, `geom_wkt` and 3 others.

---

## Quick Start

```python
from datasets import load_dataset

ds = load_dataset("electricsheepafrica/africa-ucdp-data-for-south-africa")
train = ds["train"].to_pandas()
test = ds["test"].to_pandas()

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

---

## Schema

| Column | Type | Null % | Range / Sample Values |
|---|---|---|---|
| `id` | int64 | 0.0% | 13263.0 – 551051.0 (mean 30819.6062) |
| `relid` | object | 0.0% | SAF-1989-1-516-8, SAF-1992-2-312-8, SAF-1992-2-312-300 |
| `year` | int64 | 0.0% | 1989.0 – 2024.0 (mean 1992.3119) |
| `active_year` | bool | 0.0% | |
| `code_status` | object | 0.0% | Clear |
| `type_of_violence` | int64 | 0.0% | 1.0 – 3.0 (mean 2.1755) |
| `conflict_dset_id` | int64 | 0.0% | 102.0 – 16983.0 (mean 4349.7606) |
| `conflict_new_id` | int64 | 0.0% | 298.0 – 15477.0 (mean 3950.7244) |
| `conflict_name` | object | 0.0% | Supporters of ANC - Supporters of IFP, Government of South Africa - Civilians, Supporters of IFP - Supporters of UDF |
| `dyad_dset_id` | int64 | 0.0% | 102.0 – 16983.0 (mean 4356.514) |
| `dyad_new_id` | int64 | 0.0% | 641.0 – 16983.0 (mean 4528.3606) |
| `dyad_name` | object | 0.0% | Supporters of ANC - Supporters of IFP, Government of South Africa - Civilians, Supporters of IFP - Supporters of UDF |
| `side_a_dset_id` | int64 | 0.0% | 102.0 – 8245.0 (mean 880.6877) |
| `side_a_new_id` | int64 | 0.0% | 102.0 – 8245.0 (mean 880.6877) |
| `side_a` | object | 0.0% | Supporters of ANC, Government of South Africa, Supporters of IFP |
| `side_b_dset_id` | int64 | 0.0% | 461.0 – 9999.0 (mean 2495.0967) |
| `side_b_new_id` | int64 | 0.0% | 1.0 – 8244.0 (mean 543.8463) |
| `side_b` | object | 0.0% | Supporters of IFP, Civilians, Supporters of UDF |
| `number_of_sources` | int64 | 0.0% | -1.0 – 6.0 (mean -0.8984) |
| `source_article` | object | 0.0% | The report of the Truth and Reconciliation Commission (TRC report) https://www.justice.gov.za/trc/report/, "warinangola.com/,2007-12-31,Operation Merlyn, 1989: The Nine Days War", CSVR "From Low Intensity War to Mafia War: Taxi Violence in South Africa 1987-2000", Jackie Dugard, Violence and Transition Series, vol. 4, May 2001 |
| `source_original` | object | 76.8% | police, SA, Police |
| `where_prec` | int64 | 0.0% | 1.0 – 6.0 (mean 1.277) |
| `where_coordinates` | object | 0.0% | Empangeni town, Port Shepstone town, Cape Town city |
| `where_description` | object | 0.9% | Mpumalanga town (township), Soweto town (township), Alexandra town (township) (outskirts of Johannesburg town) |
| `adm_1` | object | 0.4% | |
| `adm_2` | object | 13.7% | |
| `latitude` | float64 | 0.0% | -34.1831 – -17.3362 (mean -28.7203) |
| `longitude` | float64 | 0.0% | 13.0 – 32.3644 (mean 29.1114) |
| `geom_wkt` | object | 0.0% | |
| `priogrid_gid` | int64 | 0.0% | 80318.0 – 104794.0 (mean 88325.571) |
| `country` | object | 0.0% | |
| `iso3` | object | 0.0% | |
| `country_id` | int64 | 0.0% | 560.0 – 560.0 (mean 560.0) |
| `region` | object | 0.0% | |
| `event_clarity` | int64 | 0.0% | 1.0 – 2.0 (mean 1.0756) |
| `date_prec` | int64 | 0.0% | 1.0 – 5.0 (mean 1.362) |
| `date_start` | datetime64[ns] | 0.0% | |
| `date_end` | datetime64[ns] | 0.0% | |
| `deaths_a` | int64 | 0.0% | 0.0 – 33.0 (mean 0.4069) |
| `deaths_b` | int64 | 0.0% | |
| `deaths_civilians` | int64 | 0.0% | |
| `deaths_unknown` | int64 | 0.0% | |
| `best` | int64 | 0.0% | |
| `high` | int64 | 0.0% | |
| `low` | int64 | 0.0% | |
| `esa_source` | object | 0.0% | |
| `esa_processed` | object | 0.0% | |

---

## Numeric Summary

| Column | Min | Max | Mean | Median |
|---|---|---|---|---|
| `id` | 13263.0 | 551051.0 | 30819.6062 | 16095.0 |
| `year` | 1989.0 | 2024.0 | 1992.3119 | 1992.0 |
| `type_of_violence` | 1.0 | 3.0 | 2.1755 | 2.0 |
| `conflict_dset_id` | 102.0 | 16983.0 | 4349.7606 | 5450.0 |
| `conflict_new_id` | 298.0 | 15477.0 | 3950.7244 | 4840.0 |
| `dyad_dset_id` | 102.0 | 16983.0 | 4356.514 | 5450.0 |
| `dyad_new_id` | 641.0 | 16983.0 | 4528.3606 | 5450.0 |
| `side_a_dset_id` | 102.0 | 8245.0 | 880.6877 | 1086.0 |
| `side_a_new_id` | 102.0 | 8245.0 | 880.6877 | 1086.0 |
| `side_b_dset_id` | 461.0 | 9999.0 | 2495.0967 | 620.0 |
| `side_b_new_id` | 1.0 | 8244.0 | 543.8463 | 620.0 |
| `number_of_sources` | -1.0 | 6.0 | -0.8984 | -1.0 |
| `where_prec` | 1.0 | 6.0 | 1.277 | 1.0 |
| `latitude` | -34.1831 | -17.3362 | -28.7203 | -29.5169 |
| `longitude` | 13.0 | 32.3644 | 29.1114 | 30.3506 |

---

## 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`. 5 column(s) with >80% missing values were removed: `source_office`, `source_date`, `source_headline`, `gwnoa`, `gwnob`. 2 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 HDX 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: `source_original`.
- Refer to the [original HDX dataset page](https://data.humdata.org/dataset/ucdp-data-for-south-africa) for the publisher's own methodology notes and caveats.

---

## Citation

```bibtex
@dataset{hdx_africa_ucdp_data_for_south_africa,
title = {South Africa - Data on Conflict Events},
author = {HDX},
year = {2026},
url = {https://data.humdata.org/dataset/ucdp-data-for-south-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.*