electricsheepeurope/europe-worldbank-health

--- license: cc-by-4.0 language: - en task_categories: - tabular-classification - tabular-regression - time-series-forecasting multilinguality: monolingual size_categories: - 100K<n<1M tags: - tabular - europe - world-bank - health - wdi - development-indicators - time-series - country-level pretty_name: "Health — Europe (World Bank WDI)" --- # Health — Europe (World Bank WDI) 🇪🇺 **319,347 observations** · **44 Europe countries** · **1960–2025** · *Repackaged by [Electric Sheep Europe](https://huggingface.co/electricsheepeurope)*      ## TL;DR This dataset contains **319,347 observations** of `Health` data across **44 Europe countries**, spanning **1960–2025**, covering **235 distinct indicators**. ## About the source The World Bank's **World Development Indicators (WDI)** is the world's most-cited reference for global development data. It compiles officially-recognized international statistics from national accounts, government agencies, and partner organizations (IMF, WHO, FAO, UNESCO, ILO, UN agencies). The WDI covers over 1,400 indicators across 21 topics for 217 economies, with annual updates as new data become available. - **Source:** [World Bank](https://databank.worldbank.org/source/world-development-indicators) - **Publisher:** World Bank Group - **License:** [cc-by-4.0](https://creativecommons.org/licenses/by/4.0/) - **Topic:** Health ## Methodology Data is pulled directly from the World Bank's public API (`api.worldbank.org/v2`) for every indicator tagged to this topic across all countries in the region. Records with no observed value are dropped. Years are normalized to integers; values are coerced to float64. ## Geographic coverage 44 Europe countries · top rows shown below, sorted by row count: | Country | Rows | First year | Last year | |---------|-----:|-----------:|----------:| | `PRT` | 7,805 | 1960 | 2025 | | `ROU` | 7,770 | 1960 | 2025 | | `IRL` | 7,761 | 1960 | 2025 | | `NLD` | 7,725 | 1960 | 2025 | | `BGR` | 7,663 | 1960 | 2025 | | `MDA` | 7,663 | 1960 | 2025 | | `LTU` | 7,656 | 1960 | 2025 | | `GRC` | 7,648 | 1960 | 2025 | | `LUX` | 7,624 | 1960 | 2025 | | `ALB` | 7,579 | 1960 | 2025 | | `EST` | 7,554 | 1960 | 2025 | | `LVA` | 7,553 | 1960 | 2025 | | `MKD` | 7,546 | 1960 | 2025 | | `GBR` | 7,546 | 1960 | 2025 | | `CHE` | 7,545 | 1960 | 2025 | | ... | _29 more countries_ | | | ## Indicators (sample) - `SG.VAW.1549.ZS` — Proportion of women subjected to physical and/or sexual violence in the last 12 months (% of ever-partnered women ages 15-49) - `SG.VAW.ARGU.ZS` — Women who believe a husband is justified in beating his wife when she argues with him (%) - `SG.VAW.BURN.ZS` — Women who believe a husband is justified in beating his wife when she burns the food (%) - `SG.VAW.GOES.ZS` — Women who believe a husband is justified in beating his wife when she goes out without telling him (%) - `SG.VAW.NEGL.ZS` — Women who believe a husband is justified in beating his wife when she neglects the children (%) - `SG.VAW.REAS.ZS` — Women who believe a husband is justified in beating his wife (any of five reasons) (%) - `SG.VAW.REFU.ZS` — Women who believe a husband is justified in beating his wife when she refuses sex with him (%) - `SH.ALC.PCAP.FE.LI` — Total alcohol consumption per capita, female (liters of pure alcohol, projected estimates, female 15+ years of age) - `SH.ALC.PCAP.LI` — Total alcohol consumption per capita (liters of pure alcohol, projected estimates, 15+ years of age) - `SH.ALC.PCAP.MA.LI` — Total alcohol consumption per capita, male (liters of pure alcohol, projected estimates, male 15+ years of age) - _... and 225 more indicators_ ## Schema | Column | Type | Description | Example | |--------|------|-------------|---------| | `indicator_id` | `string` | World Bank WDI indicator code (e.g. `SP.POP.TOTL`) | `SG.VAW.GOES.ZS` | | `indicator_name` | `string` | Human-readable indicator name | `Women who believe a husband is justif…` | | `country_iso2` | `string` | ISO 3166-1 alpha-2 country code | `AL` | | `country_iso3` | `string` | ISO 3166-1 alpha-3 country code (preferred join key) | `ALB` | | `country_name` | `string` | English country name as published by the World Bank | `Albania` | | `year` | `int64` | Year of observation (CE) | `2018` | | `value` | `float64` | Observed indicator value | `3.7` | | `unit` | `string` | Unit of measurement (often empty) | `` | | `obs_status` | `string` | World Bank observation-status flag (e.g. preliminary, estimated) | `` | ## Data quality & caveats - Records with null observations are filtered out at ingest. - Indicator coverage varies by country and year; not every (indicator × country × year) combination is present. - Some countries appear under multiple aggregate groupings in WB; this dataset is filtered to **individual sovereign states only**. ## Usage ```python from datasets import load_dataset ds = load_dataset("electricsheepeurope/europe-worldbank-health") df = ds["train"].to_pandas() print(df.head()) ``` ### Filter to one country ```python germany = df[df["country_iso3"] == "DEU"] ``` ### Time-series for a single indicator ```python sample = (df[df["indicator_id"] == "SG.VAW.GOES.ZS"] .sort_values("year")) sample.plot(x="year", y="value", title="SG.VAW.GOES.ZS") ``` ### Pivot to country × year matrix ```python matrix = (df[df["indicator_id"] == "SG.VAW.GOES.ZS"] .pivot_table(index="year", columns="country_iso3", values="value")) print(matrix.tail()) ``` ## Citation ```bibtex @misc{europe_worldbank_health_2025, title = {Health — Europe (World Bank WDI)}, author = {World Bank Group}, year = {2025}, url = {https://databank.worldbank.org/source/world-development-indicators}, publisher = {HuggingFace Datasets, repackaged by Electric Sheep Europe}, howpublished = {\url{https://huggingface.co/datasets/electricsheepeurope/europe-worldbank-health}} } ``` ## License Released under [cc-by-4.0](https://creativecommons.org/licenses/by/4.0/). Original data © World Bank Group. When using this dataset, please cite both the original source above and the Electric Sheep Europe repackaging. ## About Electric Sheep Electric Sheep Europe is part of the Electric Sheep mission: a unified, ML-ready data layer for Europe on HuggingFace. We pull data from authoritative open sources, normalize the schemas, package as Parquet, and publish with consistent dataset cards so researchers and developers can use `load_dataset()` to start working in seconds. Browse the full collection: [huggingface.co/electricsheepeurope](https://huggingface.co/electricsheepeurope) --- _Provenance: ingested 2026-05-24 via the Electric Sheep pipeline. Source URL: https://databank.worldbank.org/source/world-development-indicators_This dataset contains 319,347 observations of Health data across 44 Europe countries, spanning 1960–2025, covering 235 distinct indicators. It is sourced from the World Banks World Development Indicators (WDI), pulled via public API, with null records filtered out and years/values normalized. The dataset is suitable for tabular classification, regression, and time-series forecasting tasks, focusing on health-related development indicators such as attitudes towards violence against women and alcohol consumption. Data is provided in a structured tabular format with columns including indicator ID, name, country information, year, and value.

搜集汇总

数据集介绍

构建方式

世界银行世界发展指标（WDI）作为全球发展数据的权威来源，长期整合来自国际货币基金组织、世界卫生组织、联合国等机构的官方统计。在此背景下，该数据集通过世界银行公共API（api.worldbank.org/v2）自动抓取欧洲区域所有与健康主题相关的指标数据，并剔除无观测值的记录。年份字段统一规范为整数，指标数值强制转换为float64类型，最终形成覆盖44个欧洲主权国家、1960至2025年、235项健康指标的319,347条观测记录，并以Parquet格式发布。

特点

该数据集聚焦于欧洲区域健康领域，兼具时间跨度和指标多样性。其核心特点在于涵盖235项细分健康指标，从女性暴力遭遇比例到人均酒精消费量，覆盖行为、营养、疾病、卫生系统等多维议题。地理范围限定为44个欧洲主权国家，时间序列自1960年延伸至2025年，支持跨国与长期趋势分析。数据以表格形式组织，包含指标代码、国家ISO编码、年份、数值等字段，并保留世界银行的观测状态标志，便于追溯数据质量。

使用方法

研究者可通过HuggingFace的datasets库以一行代码加载该数据集，并转换为Pandas数据框进行灵活操作。典型用法包括：按国家ISO3编码筛选特定国家子集，针对单一指标提取时间序列并排序绘制趋势图，或利用透视表将数据重塑为国家×年份矩阵以开展面板分析。这种设计兼顾了探索性分析与建模需求，适用于分类、回归及时间序列预测等任务，同时加载和转换过程无需额外数据清洗。

背景与挑战

背景概述

世界银行世界发展指标（WDI）自二十世纪后期以来始终是全球发展统计的权威基准，汇聚了来自IMF、WHO、FAO等国际组织的官方数据。2025年，Electric Sheep Europe团队将WDI中健康主题下44个欧洲国家、235项指标、319,347条观测记录从1960年至2025年的数据标准化重打包为europe-worldbank-health数据集，旨在为欧洲健康差异的跨时空研究提供可直接加载的机器学习就绪资源，显著降低公共卫生数据科学家的数据预处理负担，并推动健康不平等领域的可复现研究。

当前挑战

该数据集所面临的挑战兼具领域与工程双重维度。在领域层面，欧洲健康不平等的时空异质性要求模型能够捕捉非线性、多尺度与跨国异质效应，而现有方法对指标间复杂因果结构及缺失机制的建模仍显不足。在构建层面，WDI原始数据存在指标覆盖随国家与年份高度不均衡、零散缺失值以及部分观测状态未标注等问题，迫使管道采取过滤策略可能引入选择偏差；同时，跨国面板数据中不同国家上报口径与统计能力的差异亦为跨时空比较与模型泛化带来显著困难。

常用场景

经典使用场景

在全球健康与发展的量化研究版图中，长期跨国面板数据构成实证分析的基石。该数据集以1960至2025年、44个欧洲国家、235项健康指标、逾31万条观测的规模，为研究者提供了纵贯六十余年的欧洲健康图景。其经典使用场景涵盖面板回归、时间序列预测与跨国比较分析：研究者可提取单一指标构建国家—年份矩阵，考察健康结果的收敛或分化趋势；亦可整合多维指标，运用固定效应或动态面板模型，识别医疗支出、行为风险与健康产出之间的因果关联。

解决学术问题

健康经济学与公共卫生研究长期面临跨国数据碎片化、指标口径不一及时间跨度不足的掣肘。该数据集将世界银行WDI中分散的健康指标系统化整合，并统一为长格式结构，有效缓解了数据清洗与对齐的负担。其学术意义在于，为检验健康收敛假说、评估卫生政策干预效果、探究社会经济决定因素等核心问题提供了可复现的数据基础，推动了欧洲健康不平等、酒精消费、性别暴力等议题的跨国实证积累。

衍生相关工作

基于世界银行WDI的开放数据，学界已衍生出大量经典工作，包括全球疾病负担比较、健康支出与产出弹性估计、以及欧洲健康收敛性研究等。该数据集作为区域化重封装版本，延续了这一脉络，其衍生的相关研究涉及面板计量方法在健康领域的应用、跨国健康不平等分解，以及机器学习模型对健康指标的预测与缺失插补，为后续欧洲健康数据科学工作提供了可复用的基准。