electricsheepafrica/africa-wfp-food-prices-for-kingdom-of-eswatini

--- 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-regression - other task_ids: [] tags: - africa - humanitarian - hdx - electric-sheep-africa - economics - food-security - hxl - indicators - markets - swz pretty_name: "Kingdom of Eswatini - Food Prices" dataset_info: splits: - name: train num_examples: 4707 - name: test num_examples: 1176 --- # Kingdom of Eswatini - Food Prices **Publisher:** WFP - World Food Programme · **Source:** [HDX](https://data.humdata.org/dataset/wfp-food-prices-for-kingdom-of-eswatini) · **License:** `cc-by-igo` · **Updated:** 2023-08-20 --- ## Abstract This dataset contains Food Prices data for Eswatini, sourced from the World Food Programme Price Database. The World Food Programme Price Database covers foods such as maize, rice, beans, fish, and sugar for 98 countries and some 3000 markets. It is updated weekly but contains to a large extent monthly data. The data goes back as far as 1992 for a few countries, although many countries started reporting from 2003 or thereafter. Each row in this dataset represents subnational administrative unit observations. Temporal coverage is indicated by the `date` column(s). Geographic scope: **SWZ**. *Curated into ML-ready Parquet format by [Electric Sheep Africa](https://huggingface.co/electricsheepafrica).* --- ## Dataset Characteristics | | | |---|---| | **Domain** | Food security and nutrition | | **Unit of observation** | Subnational administrative unit observations | | **Rows (total)** | 5,884 | | **Columns** | 16 (4 numeric, 11 categorical, 1 datetime) | | **Train split** | 4,707 rows | | **Test split** | 1,176 rows | | **Geographic scope** | SWZ | | **Publisher** | WFP - World Food Programme | | **HDX last updated** | 2023-08-20 | --- ## Variables **Geographic** — `admin1` (Lubombo, Manzini, Hhohho), `admin2` (MANZINI NORTH, MAYIWANE, SITHOBELA), `latitude` (range -27.0–-26.0), `longitude` (range 31.37–31.75), `category` (cereals and tubers, miscellaneous food, pulses and nuts) and 3 others. **Temporal** — `date`. **Outcome / Measurement** — `priceflag` (actual, forecast, #item+price+flag), `price` (range 0.0–820.0), `usdprice` (range 0.0–57.5439). **Identifier / Metadata** — `esa_source` (HDX), `esa_processed`. **Other** — `market` (Manzini, National Average, Hhohho), `unit` (KG, 700 G, 10 KG). --- ## Quick Start ```python from datasets import load_dataset ds = load_dataset("electricsheepafrica/africa-wfp-food-prices-for-kingdom-of-eswatini") train = ds["train"].to_pandas() test = ds["test"].to_pandas() print(train.shape) train.head() ``` --- ## Schema | Column | Type | Null % | Range / Sample Values | |---|---|---|---| | `date` | datetime64[ns] | 0.0% | | | `admin1` | object | 20.6% | Lubombo, Manzini, Hhohho | | `admin2` | object | 20.6% | MANZINI NORTH, MAYIWANE, SITHOBELA | | `market` | object | 0.0% | Manzini, National Average, Hhohho | | `latitude` | float64 | 20.6% | -27.0 – -26.0 (mean -26.4671) | | `longitude` | float64 | 20.6% | 31.37 – 31.75 (mean 31.5671) | | `category` | object | 0.0% | cereals and tubers, miscellaneous food, pulses and nuts | | `commodity` | object | 0.0% | Rice, Sugar (brown), Beans (sugar) | | `unit` | object | 0.0% | KG, 700 G, 10 KG | | `priceflag` | object | 0.0% | actual, forecast, #item+price+flag | | `pricetype` | object | 0.0% | Retail, Wholesale, #item+price+type | | `currency` | object | 0.0% | SZL, #currency | | `price` | float64 | 0.0% | 0.0 – 820.0 (mean 20.4996) | | `usdprice` | float64 | 0.0% | 0.0 – 57.5439 (mean 1.5781) | | `esa_source` | object | 0.0% | HDX | | `esa_processed` | object | 0.0% | | --- ## Numeric Summary | Column | Min | Max | Mean | Median | |---|---|---|---|---| | `latitude` | -27.0 | -26.0 | -26.4671 | -26.48 | | `longitude` | 31.37 | 31.75 | 31.5671 | 31.5 | | `price` | 0.0 | 820.0 | 20.4996 | 11.99 | | `usdprice` | 0.0 | 57.5439 | 1.5781 | 0.8874 | --- ## 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) 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 WFP - World Food Programme 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: `admin1`, `admin2`, `latitude`, `longitude`. - Refer to the [original HDX dataset page](https://data.humdata.org/dataset/wfp-food-prices-for-kingdom-of-eswatini) for the publisher's own methodology notes and caveats. --- ## Citation ```bibtex @dataset{hdx_africa_wfp_food_prices_for_kingdom_of_eswatini, title = {Kingdom of Eswatini - Food Prices}, author = {WFP - World Food Programme}, year = {2023}, url = {https://data.humdata.org/dataset/wfp-food-prices-for-kingdom-of-eswatini}, 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: - 采集所得 language: - 英语 license: CC-BY-4.0（知识共享署名4.0国际许可协议） multilinguality: - 单语言 size_categories: - 1000 < N < 10000 source_datasets: - 原创数据集 task_categories: - 表格回归、其他 task_ids: - 无 tags: - 非洲、人道主义、HDX（Humanitarian Data Exchange）、Electric Sheep Africa、经济学、粮食安全、HXL（Humanitarian Exchange Language）、指标、市场、SWZ pretty_name: "埃斯瓦蒂尼王国 - 食品价格" dataset_info: splits: - 名称：训练集 样本数：4707 - 名称：测试集 样本数：1176 # 埃斯瓦蒂尼王国 - 食品价格 **发布方**：世界粮食计划署（WFP, World Food Programme） · **数据源**：[HDX（人道主义数据交换平台）](https://data.humdata.org/dataset/wfp-food-prices-for-kingdom-of-eswatini) · **许可证**：`cc-by-igo` · **更新时间**：2023年8月20日 --- ## 摘要 本数据集包含源自世界粮食计划署价格数据库的埃斯瓦蒂尼食品价格数据。世界粮食计划署价格数据库覆盖了98个国家约3000个市场的玉米、大米、豆类、鱼类及食糖等食品价格信息。该数据库每周更新，但其数据主体为月度统计。部分国家的最早数据可追溯至1992年，多数国家则从2003年及之后开始上报数据。 本数据集的每一行均代表次国家级行政单元的观测记录，时间覆盖范围由`date`（日期）列标识。地理覆盖范围：**SWZ（埃斯瓦蒂尼王国）**。 *本数据集已由[Electric Sheep Africa](https://huggingface.co/electricsheepafrica)整理为适合机器学习使用的Parquet格式。* --- ## 数据集特征 | | | |---|---| | **应用领域** | 粮食安全与营养 | | **观测单元** | 次国家级行政单元观测记录 | | **总样本行数** | 5884条 | | **列数** | 16列（4个数值型、11个分类型、1个日期时间型） | | **训练集划分** | 4707条 | | **测试集划分** | 1176条 | | **地理覆盖范围** | SWZ（埃斯瓦蒂尼王国） | | **发布方** | 世界粮食计划署（WFP） | | **HDX平台最后更新时间** | 2023年8月20日 | --- ## 变量说明 **地理类变量** — `admin1`（一级行政区：卢邦博、曼齐尼、霍霍），`admin2`（二级行政区：曼齐尼北区、马伊瓦内、西托贝拉），`latitude`（纬度范围：-27.0~-26.0），`longitude`（经度范围：31.37~31.75），`category`（食品类别：谷物与块根作物、杂项食品、豆类与坚果）及其他3类。 **时间类变量** — `date`（日期）。 **结果/测量类变量** — `priceflag`（价格标记：实际价格、预测价格、#item+price+flag），`price`（本币价格范围：0.0~820.0），`usdprice`（美元计价价格范围：0.0~57.5439）。 **标识符/元数据类变量** — `esa_source`（数据来源：HDX），`esa_processed`（处理标记）。 **其他变量** — `market`（市场：曼齐尼、全国平均、霍霍），`unit`（计价单位：千克、700克、10千克）。 --- ## 快速上手 python from datasets import load_dataset ds = load_dataset("electricsheepafrica/africa-wfp-food-prices-for-kingdom-of-eswatini") 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 | 范围/示例值 | |---|---|---|---| | `date` | datetime64[ns] | 0.0% | | | `admin1` | object | 20.6% | 卢邦博、曼齐尼、霍霍 | | `admin2` | object | 20.6% | 曼齐尼北区、马伊瓦内、西托贝拉 | | `market` | object | 0.0% | 曼齐尼、全国平均、霍霍 | | `latitude` | float64 | 20.6% | -27.0 ~ -26.0（均值：-26.4671） | | `longitude` | float64 | 20.6% | 31.37 ~ 31.75（均值：31.5671） | | `category` | object | 0.0% | 谷物与块根作物、杂项食品、豆类与坚果 | | `commodity` | object | 0.0% | 大米、（棕）食糖、豆类 | | `unit` | object | 0.0% | 千克、700克、10千克 | | `priceflag` | object | 0.0% | 实际价格、预测价格、#item+price+flag | | `pricetype` | object | 0.0% | 零售、批发、#item+price+type | | `currency` | object | 0.0% | 斯威士兰里兰吉尼（SZL）、#currency | | `price` | float64 | 0.0% | 0.0 ~ 820.0（均值：20.4996） | | `usdprice` | float64 | 0.0% | 0.0 ~ 57.5439（均值：1.5781） | | `esa_source` | object | 0.0% | HDX | | `esa_processed` | object | 0.0% | | --- ## 数值统计汇总 | 列名 | 最小值 | 最大值 | 均值 | 中位数 | |---|---|---|---|---| | `latitude` | -27.0 | -26.0 | -26.4671 | -26.48 | | `longitude` | 31.37 | 31.75 | 31.5671 | 31.5 | | `price` | 0.0 | 820.0 | 20.4996 | 11.99 | | `usdprice` | 0.0 | 57.5439 | 1.5781 | 0.8874 | --- ## 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为蛇形命名法。常见缺失值标记（`N/A`、`null`、`none`、`-`、`unknown`、`no data`、`#N/A`）被统一替换为`NaN`。基于解析成功率（阈值85%），将5列从字符串类型转换为数值型或日期时间型。本数据集以固定随机种子（42）按80/20比例划分为训练集与测试集，并保存为Snappy压缩的Parquet格式。 --- ## 数据集局限性 - 本数据集源自世界粮食计划署，Electric Sheep Africa未对其进行独立验证。 - 自动化清洗流程无法修正原始数据收集阶段的错报值、定义不一致或抽样偏差问题。 - 以下列的缺失率超过20%，在建模过程中需谨慎使用：`admin1`、`admin2`、`latitude`、`longitude`。 - 如需了解发布方的方法说明与免责声明，请参阅[原始HDX数据集页面](https://data.humdata.org/dataset/wfp-food-prices-for-kingdom-of-eswatini)。 --- ## 引用格式 bibtex @dataset{hdx_africa_wfp_food_prices_for_kingdom_of_eswatini, title = {埃斯瓦蒂尼王国 - 食品价格}, author = {世界粮食计划署（WFP）}, year = {2023}, url = {https://data.humdata.org/dataset/wfp-food-prices-for-kingdom-of-eswatini}, note = {由Electric Sheep Africa重新打包为机器学习可用数据集（https://huggingface.co/electricsheepafrica)} } --- *[Electric Sheep Africa](https://huggingface.co/electricsheepafrica) — 非洲机器学习数据集基础设施。尼日利亚拉各斯。*