electricsheepafrica/africa-disability-lesotho

---
annotations_creators:
- no-annotation
language_creators:
- found
language:
- en
license: other
multilinguality:
- monolingual
size_categories:
- 10K<n<100K
source_datasets:
- original
task_categories:
- tabular-classification
task_ids: []
tags:
- africa
- humanitarian
- hdx
- electric-sheep-africa
- disability
- disease
- environment
- health
- indicators
- maternity
- mental-health
- nutrition
- lso
pretty_name: "Lesotho - Health Indicators"
dataset_info:
splits:
- name: train
num_examples: 17330
- name: test
num_examples: 4332
---

# Lesotho - Health Indicators

**Publisher:** World Health Organization · **Source:** [HDX](https://data.humdata.org/dataset/who-data-for-lso) · **License:** `hdx-other` · **Updated:** 2026-04-15

---

## Abstract

This dataset contains data from WHO's [data portal](https://www.who.int/gho/en/) covering the following categories:

Air pollution, Child mortality, Dementia diagnosis, treatment and care, Environment and health, Food safety, Global Dementia Observatory (GDO), Global Health Estimates: Life expectancy and leading causes of death and disability, Global Information System on Alcohol and Health, Global Patient Safety Observatory, HIV, Health financing, Health systems, Health taxes, Health workforce, Hepatitis, Immunization coverage and vaccine-preventable diseases, Malaria, Maternal and reproductive health, Mental health, Neglected tropical diseases, Noncommunicable diseases, Nutrition, Oral Health, Priority health technologies, Resources for Substance Use Disorders, Road Safety, Sexually Transmitted Infections, Substance use disorders service coverage, Tobacco control, Tuberculosis, Universal health coverage (UHC), SDG Target 3.8, Vaccine-preventable communicable diseases, Violence prevention, Water, sanitation and hygiene (WASH), World Health Statistics.

For links to individual indicator metadata, see resource descriptions.

Each row in this dataset represents first-level administrative unit observations. Data was last updated on HDX on 2026-04-15. Geographic scope: **LSO**.

*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)** | 21,663 |
| **Columns** | 19 (6 numeric, 13 categorical, 0 datetime) |
| **Train split** | 17,330 rows |
| **Test split** | 4,332 rows |
| **Geographic scope** | LSO |
| **Publisher** | World Health Organization |
| **HDX last updated** | 2026-04-15 |

---

## Variables

**Geographic** — `gho_display` (Number of deaths in children aged <5 years, by cause, Distribution of causes of death among children aged < 5 years (%), Deaths per 1 000 live births), `year_display` (range 1952.0–2030.0), `startyear` (range 1952.0–2030.0), `endyear` (range 1952.0–2030.0), `region_code` (AFR) and 4 others.

**Outcome / Measurement** — `value`.

**Identifier / Metadata** — `gho_code` (MORT_100, MORT_300, MORT_200), `dimension_code` (SEX_BTSX, SEX_FMLE, SEX_MLE), `dimension_name` (Both sexes, Female, Male), `esa_source`, `esa_processed`.

**Other** — `gho_url` (https://www.who.int/data/gho/data/indicators/indicator-details/GHO/number-of-deaths, https://www.who.int/data/gho/data/indicators/indicator-details/GHO/distribution-of-causes-of-death-among-children-aged-5-years-%28-%29, https://www.who.int/data/gho/data/indicators/indicator-details/GHO/deaths-per-1-000-live-births), `numeric` (range -0.0625–12429001.51), `low` (range -0.2664–280000.0), `high` (range 0.0–330000.0).

---

## Quick Start

```python
from datasets import load_dataset

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

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

---

## Schema

| Column | Type | Null % | Range / Sample Values |
|---|---|---|---|
| `gho_code` | object | 0.0% | MORT_100, MORT_300, MORT_200 |
| `gho_display` | object | 0.0% | Number of deaths in children aged <5 years, by cause, Distribution of causes of death among children aged < 5 years (%), Deaths per 1 000 live births |
| `gho_url` | object | 0.0% | https://www.who.int/data/gho/data/indicators/indicator-details/GHO/number-of-deaths, https://www.who.int/data/gho/data/indicators/indicator-details/GHO/distribution-of-causes-of-death-among-children-aged-5-years-%28-%29, https://www.who.int/data/gho/data/indicators/indicator-details/GHO/deaths-per-1-000-live-births |
| `year_display` | int64 | 0.0% | 1952.0 – 2030.0 (mean 2009.3139) |
| `startyear` | int64 | 0.0% | 1952.0 – 2030.0 (mean 2009.3107) |
| `endyear` | int64 | 0.0% | 1952.0 – 2030.0 (mean 2009.3139) |
| `region_code` | object | 0.0% | AFR |
| `region_display` | object | 0.0% | Africa |
| `country_code` | object | 0.0% | LSO |
| `country_display` | object | 0.0% | Lesotho |
| `dimension_type` | object | 9.1% | SEX, RESIDENCEAREATYPE, AGEGROUP |
| `dimension_code` | object | 9.1% | SEX_BTSX, SEX_FMLE, SEX_MLE |
| `dimension_name` | object | 9.1% | Both sexes, Female, Male |
| `numeric` | float64 | 9.9% | -0.0625 – 12429001.51 (mean 3248.4418) |
| `value` | object | 0.1% | |
| `low` | float64 | 32.3% | -0.2664 – 280000.0 (mean 1529.0957) |
| `high` | float64 | 32.3% | 0.0 – 330000.0 (mean 2406.0427) |
| `esa_source` | object | 0.0% | |
| `esa_processed` | object | 0.0% | |

---

## Numeric Summary

| Column | Min | Max | Mean | Median |
|---|---|---|---|---|
| `year_display` | 1952.0 | 2030.0 | 2009.3139 | 2011.0 |
| `startyear` | 1952.0 | 2030.0 | 2009.3107 | 2011.0 |
| `endyear` | 1952.0 | 2030.0 | 2009.3139 | 2011.0 |
| `numeric` | -0.0625 | 12429001.51 | 3248.4418 | 16.8951 |
| `low` | -0.2664 | 280000.0 | 1529.0957 | 13.3 |
| `high` | 0.0 | 330000.0 | 2406.0427 | 25.735 |

---

## 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`. 750 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 World Health Organization 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: `low`, `high`.
- Refer to the [original HDX dataset page](https://data.humdata.org/dataset/who-data-for-lso) for the publisher's own methodology notes and caveats.

---

## Citation

```bibtex
@dataset{hdx_africa_disability_lesotho,
title = {Lesotho - Health Indicators},
author = {World Health Organization},
year = {2026},
url = {https://data.humdata.org/dataset/who-data-for-lso},
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:
- 采集自现有数据
language:
- 英语
license:
- 其他
multilinguality:
- 单语言
size_categories:
- 10K<n<100K
source_datasets:
- 原始数据集
task_categories:
- 表格分类（tabular-classification）
task_ids: []
tags:
- 非洲
- 人道主义
- HDX（Humanitarian Data Exchange，人道主义数据交换平台）
- electric-sheep-africa
- 残疾
- 疾病
- 环境
- 健康
- 指标
- 孕产健康
- 心理健康
- 营养
- LSO（莱索托国家代码）
pretty_name: "莱索托——健康指标"
dataset_info:
splits:
- name: train
num_examples: 17330
- name: test
num_examples: 4332

# 莱索托——健康指标

**发布方：世界卫生组织（World Health Organization, WHO） · 来源：** [HDX（Humanitarian Data Exchange，人道主义数据交换平台）](https://data.humdata.org/dataset/who-data-for-lso) · **许可协议：** `hdx-other` · **更新时间：** 2026-04-15

---

## 摘要
本数据集收录源自世界卫生组织（WHO）[数据门户](https://www.who.int/gho/en/)的相关数据，涵盖以下类别：
空气污染、儿童死亡率、痴呆症诊断、治疗与照护、环境与健康、食品安全、全球痴呆症观测站（Global Dementia Observatory, GDO）、全球健康估算：预期寿命及主要死亡与残疾原因、全球酒精与健康信息系统、全球患者安全观测站、艾滋病、卫生筹资、卫生系统、健康税、卫生人力、肝炎、免疫接种覆盖与疫苗可预防疾病、疟疾、孕产妇与生殖健康、心理健康、被忽视的热带病、非传染性疾病、营养、口腔健康、优先健康技术、物质使用障碍相关资源、道路安全、性传播感染、物质使用障碍服务覆盖、烟草控制、结核病、全民健康覆盖（Universal Health Coverage, UHC）、可持续发展目标3.8指标、疫苗可预防传染病、暴力预防、水、环境卫生与个人卫生（Water, Sanitation and Hygiene, WASH）、世界卫生统计。

如需获取单个指标元数据的链接，请参阅资源描述文档。

本数据集的每一行均代表一级行政单元的观测数据。本数据集最后一次在HDX平台更新的时间为2026-04-15，地理覆盖范围为**莱索托（LSO）**。

*本数据集已由[Electric Sheep Africa（电动绵羊非洲团队）](https://huggingface.co/electricsheepafrica)整理为适配机器学习的Parquet格式。*

---

## 数据集特征
| 类别 | 详情 |
|---|---|
| **研究领域** | 粮食安全与营养 |
| **观测单元** | 一级行政单元观测数据 |
| **总数据行数** | 21,663 |
| **总列数** | 19列（6列数值型、13列分类型、0列日期型） |
| **训练集划分** | 17,330行 |
| **测试集划分** | 4,332行 |
| **地理覆盖范围** | 莱索托（LSO） |
| **发布方** | 世界卫生组织 |
| **HDX平台最后更新时间** | 2026-04-15 |

---

## 变量说明
**地理相关变量** — `gho_display`（按病因划分的5岁以下儿童死亡数、5岁以下儿童死亡原因分布占比、每1000活产儿死亡数）、`year_display`（取值范围1952.0–2030.0）、`startyear`（取值范围1952.0–2030.0）、`endyear`（取值范围1952.0–2030.0）、`region_code`（AFR，非洲区域代码）及另外4个变量。

**结果与测量变量** — `value`。

**标识与元数据变量** — `gho_code`（MORT_100、MORT_300、MORT_200等指标编码）、`dimension_code`（SEX_BTSX、SEX_FMLE、SEX_MLE等维度编码）、`dimension_name`（合计性别、女性、男性等维度名称）、`esa_source`、`esa_processed`。

**其他变量** — `gho_url`（指标详情链接）、`numeric`（取值范围-0.0625–12429001.51）、`low`（置信下限，取值范围-0.2664–280000.0）、`high`（置信上限，取值范围0.0–330000.0）。

---

## 快速上手
python
from datasets import load_dataset

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

print(train.shape)
train.head()

---

## 数据Schema
| 列名 | 数据类型 | 缺失率 | 取值范围/示例值 |
|---|---|---|---|
| `gho_code` | 字符型（object） | 0.0% | MORT_100、MORT_300、MORT_200 |
| `gho_display` | 字符型（object） | 0.0% | 按病因划分的5岁以下儿童死亡数、5岁以下儿童死亡原因分布占比、每1000活产儿死亡数 |
| `gho_url` | 字符型（object） | 0.0% | 世界卫生组织指标详情链接 |
| `year_display` | 64位整数型（int64） | 0.0% | 1952.0 – 2030.0（均值2009.3139） |
| `startyear` | 64位整数型（int64） | 0.0% | 1952.0 – 2030.0（均值2009.3107） |
| `endyear` | 64位整数型（int64） | 0.0% | 1952.0 – 2030.0（均值2009.3139） |
| `region_code` | 字符型（object） | 0.0% | AFR（非洲区域代码） |
| `region_display` | 字符型（object） | 0.0% | 非洲 |
| `country_code` | 字符型（object） | 0.0% | LSO（莱索托国家代码） |
| `country_display` | 字符型（object） | 0.0% | 莱索托 |
| `dimension_type` | 字符型（object） | 9.1% | 性别、居住区域类型、年龄组 |
| `dimension_code` | 字符型（object） | 9.1% | SEX_BTSX、SEX_FMLE、SEX_MLE |
| `dimension_name` | 字符型（object） | 9.1% | 合计性别、女性、男性 |
| `numeric` | 64位浮点型（float64） | 9.9% | -0.0625 – 12429001.51（均值3248.4418） |
| `value` | 字符型（object） | 0.1% | 无 |
| `low` | 64位浮点型（float64） | 32.3% | -0.2664 – 280000.0（均值1529.0957） |
| `high` | 64位浮点型（float64） | 32.3% | 0.0 – 330000.0（均值2406.0427） |
| `esa_source` | 字符型（object） | 0.0% | 无 |
| `esa_processed` | 字符型（object） | 0.0% | 无 |

---

## 数值型变量统计摘要
| 列名 | 最小值 | 最大值 | 均值 | 中位数 |
|---|---|---|---|---|
| `year_display` | 1952.0 | 2030.0 | 2009.3139 | 2011.0 |
| `startyear` | 1952.0 | 2030.0 | 2009.3107 | 2011.0 |
| `endyear` | 1952.0 | 2030.0 | 2009.3139 | 2011.0 |
| `numeric` | -0.0625 | 12429001.51 | 3248.4418 | 16.8951 |
| `low` | -0.2664 | 280000.0 | 1529.0957 | 13.3 |
| `high` | 0.0 | 330000.0 | 2406.0427 | 25.735 |

---

## 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。所有列名均转换为小写，并统一规范为蛇形命名法（snake_case）。常见的缺失值标记（`N/A`、`null`、`none`、`-`、`unknown`、`no data`、`#N/A`）被统一替换为`NaN`。共移除750条完全重复的数据行。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，并以Snappy压缩的Parquet格式存储。

---

## 数据集局限性
1. 本数据集源自世界卫生组织，并未经Electric Sheep Africa（电动绵羊非洲团队）独立验证。
2. 自动化数据清洗无法修正原始数据集中的错报值、定义不一致问题或抽样偏差。
3. 以下列的缺失率超过20%，在建模过程中需谨慎使用：`low`（置信下限）、`high`（置信上限）。
4. 如需查看发布方提供的方法学说明与免责声明，请参阅[原始HDX数据集页面](https://data.humdata.org/dataset/who-data-for-lso)。

---

## 引用格式
bibtex
@dataset{hdx_africa_disability_lesotho,
title = {Lesotho - Health Indicators},
author = {World Health Organization},
year = {2026},
url = {https://data.humdata.org/dataset/who-data-for-lso},
note = {Repackaged for machine learning by Electric Sheep Africa (https://huggingface.co/electricsheepafrica)}
}

---

*[Electric Sheep Africa（电动绵羊非洲团队）](https://huggingface.co/electricsheepafrica) — 非洲机器学习数据集基础设施。尼日利亚拉各斯。*