electricsheepafrica/africa-ner-views-conflict-forecasts

---
annotations_creators:
- no-annotation
language_creators:
- found
language:
- en
license: cc-by-sa-4.0
multilinguality:
- monolingual
size_categories:
- n<1K
source_datasets:
- original
task_categories:
- tabular-classification
- tabular-regression
task_ids: []
tags:
- africa
- humanitarian
- hdx
- electric-sheep-africa
- conflict-violence
- fatalities
- forecasting
- hxl
- ner
pretty_name: "Niger - VIEWS conflict forecasts"
dataset_info:
splits:
- name: train
num_examples: 28
- name: test
num_examples: 7
---

# Niger - VIEWS conflict forecasts

**Publisher:** Violence & Impacts Early-Warning System · **Source:** [HDX](https://data.humdata.org/dataset/ner-views-conflict-forecasts) · **License:** `cc-by-sa` · **Updated:** 2026-04-01

---

## Abstract

The Violence & Impacts Early-Warning System (VIEWS) is an award-winning conflict prediction system that generates monthly forecasts for violent conflicts across the world up to three years in advance. It is supported by the iterative research and development activities undertaken by the VIEWS consortium.

Each row in this dataset represents country-level aggregates. Data was last updated on HDX on 2026-04-01. Geographic scope: **NER**.

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

---

## Dataset Characteristics

| | |
|---|---|
| **Domain** | Conflict and security |
| **Unit of observation** | Country-level aggregates |
| **Rows (total)** | 36 |
| **Columns** | 12 (8 numeric, 4 categorical, 0 datetime) |
| **Train split** | 28 rows |
| **Test split** | 7 rows |
| **Geographic scope** | NER |
| **Publisher** | Violence & Impacts Early-Warning System |
| **HDX last updated** | 2026-04-01 |

---

## Variables

**Geographic** — `country_id` (range 78.0–78.0), `isoab` (NER), `year` (range 2026.0–2029.0).

**Temporal** — `month_id` (range 555.0–590.0), `month` (range 1.0–12.0).

**Identifier / Metadata** — `name` (Niger), `gwcode` (range 436.0–436.0), `esa_source` (HDX), `esa_processed` (2026-04-06).

**Other** — `main_mean_ln` (range 3.7067–4.0223), `main_mean` (range 39.7188–54.8282), `main_dich` (range 0.9821–0.9987).

---

## Quick Start

```python
from datasets import load_dataset

ds = load_dataset("electricsheepafrica/africa-ner-views-conflict-forecasts")
train = ds["train"].to_pandas()
test = ds["test"].to_pandas()

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

---

## Schema

| Column | Type | Null % | Range / Sample Values |
|---|---|---|---|
| `country_id` | int64 | 0.0% | 78.0 – 78.0 (mean 78.0) |
| `month_id` | int64 | 0.0% | 555.0 – 590.0 (mean 572.5) |
| `name` | object | 0.0% | Niger |
| `gwcode` | int64 | 0.0% | 436.0 – 436.0 (mean 436.0) |
| `isoab` | object | 0.0% | NER |
| `year` | int64 | 0.0% | 2026.0 – 2029.0 (mean 2027.1667) |
| `month` | int64 | 0.0% | 1.0 – 12.0 (mean 6.5) |
| `main_mean_ln` | float64 | 0.0% | 3.7067 – 4.0223 (mean 3.8467) |
| `main_mean` | float64 | 0.0% | 39.7188 – 54.8282 (mean 46.0149) |
| `main_dich` | float64 | 0.0% | 0.9821 – 0.9987 (mean 0.993) |
| `esa_source` | object | 0.0% | HDX |
| `esa_processed` | object | 0.0% | 2026-04-06 |

---

## Numeric Summary

| Column | Min | Max | Mean | Median |
|---|---|---|---|---|
| `country_id` | 78.0 | 78.0 | 78.0 | 78.0 |
| `month_id` | 555.0 | 590.0 | 572.5 | 572.5 |
| `gwcode` | 436.0 | 436.0 | 436.0 | 436.0 |
| `year` | 2026.0 | 2029.0 | 2027.1667 | 2027.0 |
| `month` | 1.0 | 12.0 | 6.5 | 6.5 |
| `main_mean_ln` | 3.7067 | 4.0223 | 3.8467 | 3.8351 |
| `main_mean` | 39.7188 | 54.8282 | 46.0149 | 45.297 |
| `main_dich` | 0.9821 | 0.9987 | 0.993 | 0.9939 |

---

## 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`. 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 Violence & Impacts Early-Warning System 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](https://data.humdata.org/dataset/ner-views-conflict-forecasts) for the publisher's own methodology notes and caveats.

---

## Citation

```bibtex
@dataset{hdx_africa_ner_views_conflict_forecasts,
title = {Niger - VIEWS conflict forecasts},
author = {Violence & Impacts Early-Warning System},
year = {2026},
url = {https://data.humdata.org/dataset/ner-views-conflict-forecasts},
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.*

annotations_creators:
- 无标注
language_creators:
- 公开数据源采集
language:
- 英语
license: cc-by-sa-4.0
multilinguality:
- 单语种
size_categories:
- 样本量小于1000
source_datasets:
- 原始数据集
task_categories:
- 表格分类
- 表格回归
task_ids: []
tags:
- 非洲
- 人道主义
- HDX（Humanitarian Data Exchange）
- Electric Sheep Africa
- 冲突暴力
- 伤亡人数
- 预测
- HXL（Humanitarian Exchange Language）
- 命名实体识别（Named Entity Recognition，简称NER）
pretty_name: "尼日尔——VIEWS冲突预测数据集"
dataset_info:
splits:
- name: 训练集
num_examples: 28
- name: 测试集
num_examples: 7

# 尼日尔——VIEWS冲突预测数据集

**发布方**：暴力与影响早期预警系统（Violence & Impacts Early-Warning System，简称VIEWS） · **来源**：[HDX（Humanitarian Data Exchange）](https://data.humdata.org/dataset/ner-views-conflict-forecasts) · **许可协议**：`cc-by-sa` · **最后更新时间**：2026-04-01

---

## 摘要

暴力与影响早期预警系统（VIEWS）是一款屡获殊荣的全球暴力冲突月度预测系统，可提前三年生成全球各地暴力冲突的月度预测结果，其研发依托VIEWS联盟开展的迭代研究与开发活动。

本数据集的每一行均代表国家级聚合数据。数据最后一次在HDX平台更新的时间为2026-04-01。地理覆盖范围：**NER（尼日尔）**。

*本数据集已由[Electric Sheep Africa](https://huggingface.co/electricsheepafrica)整理为可供机器学习直接使用的Parquet格式文件。*

---

## 数据集特征

| | |
|---|---|
| **所属领域** | 冲突与安全 |
| **观测单元** | 国家级聚合数据 |
| **总行数** | 36 |
| **列数** | 12（其中8列为数值型，4列为分类型，无日期时间型列） |
| **训练集样本量** | 28 |
| **测试集样本量** |7 |
| **地理覆盖范围** | NER（尼日尔） |
| **发布方** | 暴力与影响早期预警系统 |
| **HDX平台最后更新时间** | 2026-04-01 |

---

## 变量说明

**地理类变量**：`country_id`（取值范围78.0–78.0）、`isoab`（代码为NER）、`year`（取值范围2026.0–2029.0）。
**时间类变量**：`month_id`（取值范围555.0–590.0）、`month`（取值范围1.0–12.0）。
**标识符/元数据变量**：`name`（值为"尼日尔"）、`gwcode`（取值范围436.0–436.0）、`esa_source`（值为HDX）、`esa_processed`（处理时间为2026-04-06）。
**其他变量**：`main_mean_ln`（取值范围3.7067–4.0223）、`main_mean`（取值范围39.7188–54.8282）、`main_dich`（取值范围0.9821–0.9987）。

---

## 快速上手

python
from datasets import load_dataset

ds = load_dataset("electricsheepafrica/africa-ner-views-conflict-forecasts")
train = ds["train"].to_pandas()
test = ds["test"].to_pandas()

print(train.shape)
train.head()

---

## 数据模式

| 列名 | 数据类型 | 缺失率 | 取值范围/示例值 |
|---|---|---|---|
| `country_id` | int64 | 0.0% | 78.0 – 78.0（均值为78.0） |
| `month_id` | int64 | 0.0% | 555.0 – 590.0（均值为572.5） |
| `name` | object | 0.0% | "尼日尔" |
| `gwcode` | int64 | 0.0% | 436.0 – 436.0（均值为436.0） |
| `isoab` | object | 0.0% | NER |
| `year` | int64 | 0.0% | 2026.0 – 2029.0（均值为2027.1667） |
| `month` | int64 | 0.0% | 1.0 – 12.0（均值为6.5） |
| `main_mean_ln` | float64 | 0.0% | 3.7067 – 4.0223（均值为3.8467） |
| `main_mean` | float64 | 0.0% | 39.7188 – 54.8282（均值为46.0149） |
| `main_dich` | float64 | 0.0% | 0.9821 – 0.9987（均值为0.993） |
| `esa_source` | object | 0.0% | HDX |
| `esa_processed` | object | 0.0% | 2026-04-06 |

---

## 数值型变量统计摘要

| 列名 | 最小值 | 最大值 | 均值 | 中位数 |
|---|---|---|---|---|
| `country_id` | 78.0 | 78.0 | 78.0 | 78.0 |
| `month_id` | 555.0 | 590.0 | 572.5 | 572.5 |
| `gwcode` | 436.0 | 436.0 | 436.0 | 436.0 |
| `year` | 2026.0 | 2029.0 | 2027.1667 | 2027.0 |
| `month` | 1.0 | 12.0 | 6.5 | 6.5 |
| `main_mean_ln` | 3.7067 | 4.0223 | 3.8467 | 3.8351 |
| `main_mean` | 39.7188 | 54.8282 | 46.0149 | 45.297 |
| `main_dich` | 0.9821 | 0.9987 | 0.993 | 0.9939 |

---

## 数据整理流程

原始数据通过CKAN API从HDX平台下载，并转换为Parquet格式。列名已统一转换为小写蛇形命名法（snake_case）。常见缺失值标记（`N/A`、`null`、`none`、`-`、`unknown`、`no data`、`#N/A`）已统一替换为`NaN`。本数据集以固定随机种子（42）按照80/20的比例划分为训练集与测试集，并保存为采用Snappy压缩的Parquet格式文件。

---

## 数据集局限性

- 数据源自暴力与影响早期预警系统，尚未经过Electric Sheep Africa的独立验证。
- 自动化清洗流程无法修正原始数据收集中存在的错报值、定义不一致或采样偏差问题。
- 如需了解发布方提供的方法学说明与注意事项，请参阅[原始HDX数据集页面](https://data.humdata.org/dataset/ner-views-conflict-forecasts)。

---

## 引用格式

bibtex
@dataset{hdx_africa_ner_views_conflict_forecasts,
title = {Niger - VIEWS conflict forecasts},
author = {Violence & Impacts Early-Warning System},
year = {2026},
url = {https://data.humdata.org/dataset/ner-views-conflict-forecasts},
note = {Repackaged for machine learning by Electric Sheep Africa (https://huggingface.co/electricsheepafrica)}
}

---

*[Electric Sheep Africa](https://huggingface.co/electricsheepafrica) — 非洲地区机器学习数据集基础设施。尼日利亚拉各斯。*