| --- |
| pretty_name: China Population Open Data |
| language: |
| - zh |
| tags: |
| - tabular |
| - timeseries |
| - datasets |
| - pandas |
| - mlcroissant |
| configs: |
| - config_name: provincial_annual |
| data_files: |
| - split: train |
| path: data/provincial_annual/provincial_annual.parquet |
| - config_name: national_structure |
| data_files: |
| - split: train |
| path: data/national_structure/national_structure.parquet |
| - config_name: key_city_annual |
| data_files: |
| - split: train |
| path: data/key_city_annual/key_city_annual.parquet |
| --- |
| |
| # China Population Open Data |
|
|
| 公开版人口数据仓,面向数据分析、城市研究、可视化展示与教学复现。当前版本由 `Population` 代码仓中的构建脚本生成,目标发布到: |
|
|
| - GitHub: https://github.com/Subat-01/china-population-open-data |
| - Hugging Face: https://huggingface.co/datasets/Subat01/china-population-open-data |
|
|
| ## 中文说明 |
|
|
| ### 1. 数据内容 |
|
|
| - `provincial_annual` |
| - 31 个省级/直辖市人口年度指标。 |
| - 当前覆盖 `2011-2025`,其中 `2025` 仅纳入已补齐来源的地区。 |
| - 指标包括:`resident_total_population`、`birth_rate`、`death_rate`、`natural_growth_rate`。 |
| - `national_structure` |
| - 全国年度人口总量、城乡结构、性别结构、年龄结构、抚养比以及全国年度人口率。 |
| - 当前覆盖 `2011-2025`;年龄结构与抚养比在官方表缺失年份保留为空,不做插值。 |
| - `key_city_annual` |
| - 重点地市常住总人口公开子集。 |
| - 当前收录 12 个城市:南京市, 宁波市, 广州市, 成都市, 杭州市, 武汉市, 深圳市, 苏州市, 西安市, 郑州市, 长沙市, 青岛市。 |
|
|
| ### 2. 来源与口径 |
|
|
| - 优先使用国家统计局、地方统计局、地方政府公开统计公报、人口普查公报。 |
| - `verification_status=verified` 表示已在记录中绑定具体官方页面链接。 |
| - `verification_status=derived_from_official_table` 表示记录来自仓库缓存的官方表格快照或历史官方整理结果;这类记录的 `source_publish_date` 可能是快照/整理日期,而不是官方页面发布日期。 |
| - 绝对人口统一使用 `wan_person`,人口率使用 `permillage`,抚养比使用 `percent`。 |
|
|
| ### 3. 质量控制 |
|
|
| - 已重建 `31 省 * 14 年 * 4 指标` 的 V1 基线。 |
| - 基线校验结果:`baseline_count=1736`,`rebuilt_count=1736`,`missing_key_count=0`,`extra_key_count=0`。 |
| - 重点地市覆盖与是否建议预测见 `manifests/coverage_manifest.csv`。 |
|
|
| ### 4. 使用方式 |
|
|
| ```python |
| import pandas as pd |
| |
| provincial = pd.read_parquet("data/provincial_annual/provincial_annual.parquet") |
| national = pd.read_parquet("data/national_structure/national_structure.parquet") |
| key_city = pd.read_parquet("data/key_city_annual/key_city_annual.parquet") |
| ``` |
|
|
| ### 5. 注意事项 |
|
|
| - 本仓库不公开 `docs/` 下原始 Excel,仅公开标准化结果与追溯清单。 |
| - 本仓库采用保守来源声明策略,未承诺强开放许可证;下游使用需遵守原始官方来源条款。 |
| - 如需判断数据是否适合分析或预测,请优先查看 `coverage_manifest.csv`。 |
|
|
| ## English Summary |
|
|
| This repository packages normalized China population data for analysis and reproducible demos. |
|
|
| - `provincial_annual`: provincial / municipality annual indicators |
| - `national_structure`: national annual population structure and annual rates |
| - `key_city_annual`: selected prefecture-level city resident population series |
|
|
| Please review `DATA_USAGE_NOTICE.md` before redistribution or reuse. |
|
|