electricsheepafrica/africa-fewsnet-staple-food-price-data-for-south-sudan-weekly-6857

---
annotations_creators:
- no-annotation
language_creators:
- found
language:
- en
license: cc-by-4.0
multilinguality:
- monolingual
size_categories:
- 10K<n<100K
source_datasets:
- original
task_categories:
- tabular-regression
- other
task_ids: []
tags:
- africa
- humanitarian
- hdx
- electric-sheep-africa
- eastern-africa
- economics
- food-security
- indicators
- markets
- ssd
pretty_name: "South Sudan Weekly FEWS NET Staple Food Price Data"
dataset_info:
splits:
- name: train
num_examples: 38721
- name: test
num_examples: 9680
---

# South Sudan Weekly FEWS NET Staple Food Price Data

**Publisher:** FEWS NET · **Source:** [HDX](https://data.humdata.org/dataset/fewsnet_staple_food_price_data_for_south_sudan_weekly_6857) · **License:** `cc-by` · **Updated:** 2026-04-07

---

## Abstract

South Sudan Weekly staple food price data collected by FEWS NET since 2021.

Each row in this dataset represents country-level aggregates. Temporal coverage is indicated by the `period_date` column(s). Geographic scope: **SSD**.

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

---

## Dataset Characteristics

| | |
|---|---|
| **Domain** | Food security and nutrition |
| **Unit of observation** | Country-level aggregates |
| **Rows (total)** | 48,402 |
| **Columns** | 17 (3 numeric, 13 categorical, 1 datetime) |
| **Train split** | 38,721 rows |
| **Test split** | 9,680 rows |
| **Geographic scope** | SSD |
| **Publisher** | FEWS NET |
| **HDX last updated** | 2026-04-07 |

---

## Variables

**Geographic** — `country` (South Sudan), `admin_1` (Jonglei, Upper Nile, Central Equatoria), `longitude` (range 27.3979–33.9249), `latitude` (range 4.0928–9.8874), `price_type` (Retail, Wholesale, Wage) and 2 others.

**Temporal** — `period_date`.

**Outcome / Measurement** — `value` (range 1.0–1120000.0).

**Identifier / Metadata** — `source_document` (Famine Early Warning Systems Network (FEWS NET), South Sudan, Price), `product_source` (Local, Import), `esa_source`, `esa_processed`.

**Other** — `market` (Leer, Maiwut, Malakal, Aburoc), `cpcv2` (R01122AC, R01142AC, R01142AH), `product` (Maize Grain (White), Sorghum (Red), Sorghum (Feterita)), `unit` (kg, 3.5_kg, ea).

---

## Quick Start

```python
from datasets import load_dataset

ds = load_dataset("electricsheepafrica/africa-fewsnet-staple-food-price-data-for-south-sudan-weekly-6857")
train = ds["train"].to_pandas()
test = ds["test"].to_pandas()

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

---

## Schema

| Column | Type | Null % | Range / Sample Values |
|---|---|---|---|
| `country` | object | 0.0% | South Sudan |
| `market` | object | 0.0% | Leer, Maiwut, Malakal, Aburoc |
| `admin_1` | object | 0.0% | Jonglei, Upper Nile, Central Equatoria |
| `longitude` | float64 | 0.0% | 27.3979 – 33.9249 (mean 31.3085) |
| `latitude` | float64 | 0.0% | 4.0928 – 9.8874 (mean 7.1659) |
| `cpcv2` | object | 0.0% | R01122AC, R01142AC, R01142AH |
| `product` | object | 0.0% | Maize Grain (White), Sorghum (Red), Sorghum (Feterita) |
| `source_document` | object | 0.0% | Famine Early Warning Systems Network (FEWS NET), South Sudan, Price |
| `period_date` | datetime64[ns] | 0.0% | |
| `price_type` | object | 0.0% | Retail, Wholesale, Wage |
| `product_source` | object | 0.0% | Local, Import |
| `unit` | object | 0.0% | kg, 3.5_kg, ea |
| `unit_type` | object | 0.0% | Weight, Item, Volume |
| `currency` | object | 0.0% | |
| `value` | float64 | 54.2% | 1.0 – 1120000.0 (mean 24287.4641) |
| `esa_source` | object | 0.0% | |
| `esa_processed` | object | 0.0% | |

---

## Numeric Summary

| Column | Min | Max | Mean | Median |
|---|---|---|---|---|
| `longitude` | 27.3979 | 33.9249 | 31.3085 | 31.5547 |
| `latitude` | 4.0928 | 9.8874 | 7.1659 | 7.457 |
| `value` | 1.0 | 1120000.0 | 24287.4641 | 3920.0 |

---

## 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`. 1 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 FEWS NET 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: `value`.
- Refer to the [original HDX dataset page](https://data.humdata.org/dataset/fewsnet_staple_food_price_data_for_south_sudan_weekly_6857) for the publisher's own methodology notes and caveats.

---

## Citation

```bibtex
@dataset{hdx_africa_fewsnet_staple_food_price_data_for_south_sudan_weekly_6857,
title = {South Sudan Weekly FEWS NET Staple Food Price Data},
author = {FEWS NET},
year = {2026},
url = {https://data.humdata.org/dataset/fewsnet_staple_food_price_data_for_south_sudan_weekly_6857},
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:
- 无注释
language_creators:
- 公开源采集
language:
- 英语
license: CC BY 4.0
multilinguality:
- 单语言
size_categories:
- 10000 < 样本数 < 100000
source_datasets:
- 原创数据集
task_categories:
- 表格回归
- 其他
task_ids: []
tags:
- 非洲
- 人道主义
- HDX（人道主义数据交换，Humanitarian Data Exchange）
- 非洲电羊（Electric Sheep Africa）
- 东非
- 经济学
- 粮食安全
- 指标
- 市场
- SSD（南苏丹国家代码）
pretty_name: "南苏丹每周FEWS NET（饥荒早期预警系统网络，Famine Early Warning Systems Network）主粮价格数据集"
dataset_info:
splits:
- name: train
num_examples: 38721
- name: test
num_examples: 9680

---

# 南苏丹每周FEWS NET主粮价格数据集

**发布者：** FEWS NET · **来源：** [HDX（人道主义数据交换，Humanitarian Data Exchange）](https://data.humdata.org/dataset/fewsnet_staple_food_price_data_for_south_sudan_weekly_6857) · **许可协议：** `cc-by` · **最后更新：** 2026-04-07

---

## 摘要
本数据集由FEWS NET（饥荒早期预警系统网络）自2021年起收集的南苏丹每周主粮价格数据。数据集中每一行代表国家级汇总统计结果，时间覆盖范围由`period_date`列标识，地理覆盖范围为**SSD（南苏丹国家代码）**。本数据集已由[非洲电羊（Electric Sheep Africa）](https://huggingface.co/electricsheepafrica)整理为适配机器学习的Parquet格式。

---

## 数据集特征

| | |
|---|---|
| **研究领域** | 粮食安全与营养 |
| **观测单元** | 国家级汇总统计数据 |
| **总数据行数** | 48,402 |
| **总列数** | 17列（3个数值型列、13个分类型列、1个日期时间型列） |
| **训练集样本量** | 38,721 |
| **测试集样本量** | 9,680 |
| **地理覆盖范围** | SSD（南苏丹） |
| **发布机构** | FEWS NET |
| **HDX平台最后更新时间** | 2026-04-07 |

---

## 变量说明

**地理类变量**：`country`（南苏丹）、`admin_1`（琼莱州、上尼罗河州、中赤道州）、`longitude`（经度范围：27.3979–33.9249）、`latitude`（纬度范围：4.0928–9.8874）、`price_type`（零售、批发、薪资）及另外2个变量。
**时间类变量**：`period_date`（报告日期）。
**结果/测量变量**：`value`（数值范围：1.0–1120000.0）。
**标识符/元数据变量**：`source_document`（饥荒早期预警系统网络（FEWS NET）、南苏丹、价格数据）、`product_source`（本地、进口）、`esa_source`、`esa_processed`。
**其他变量**：`market`（里尔、迈武特、马拉卡尔、阿布罗克）、`cpcv2`（R01122AC、R01142AC、R01142AH）、`product`（白玉米粒、红高粱、费特里塔高粱）、`unit`（千克、3.5千克、件）。

---

## 快速上手

python
from datasets import load_dataset

# 加载目标数据集
ds = load_dataset("electricsheepafrica/africa-fewsnet-staple-food-price-data-for-south-sudan-weekly-6857")
# 将训练集与测试集转换为Pandas DataFrame格式
train = ds["train"].to_pandas()
test = ds["test"].to_pandas()

# 打印训练集的形状信息
print(train.shape)
# 查看训练集前5条数据
train.head()

---

## 数据 Schema

| 列名 | 数据类型 | 缺失率 | 取值范围/示例值 |
|---|---|---|---|
| `country` | 字符串（object） | 0.0% | 南苏丹 |
| `market` | 字符串（object） | 0.0% | 里尔、迈武特、马拉卡尔、阿布罗克 |
| `admin_1` | 字符串（object） | 0.0% | 琼莱州、上尼罗河州、中赤道州 |
| `longitude` | 浮点型（float64） | 0.0% | 27.3979 – 33.9249（均值：31.3085） |
| `latitude` | 浮点型（float64） | 0.0% | 4.0928 – 9.8874（均值：7.1659） |
| `cpcv2` | 字符串（object） | 0.0% | R01122AC、R01142AC、R01142AH |
| `product` | 字符串（object） | 0.0% | 白玉米粒、红高粱、费特里塔高粱 |
| `source_document` | 字符串（object） | 0.0% | 饥荒早期预警系统网络（FEWS NET）、南苏丹、价格数据 |
| `period_date` | 日期时间型（datetime64[ns]） | 0.0% | 无 |
| `price_type` | 字符串（object） | 0.0% | 零售、批发、薪资 |
| `product_source` | 字符串（object） | 0.0% | 本地、进口 |
| `unit` | 字符串（object） | 0.0% | 千克、3.5_kg、件 |
| `unit_type` | 字符串（object） | 0.0% | 重量、物品、体积 |
| `currency` | 字符串（object） | 0.0% | 无 |
| `value` | 浮点型（float64） | 54.2% | 1.0 – 1120000.0（均值：24287.4641） |
| `esa_source` | 字符串（object） | 0.0% | 无 |
| `esa_processed` | 字符串（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 |
|---|---|---|---|---|
| `longitude` | 27.3979 | 33.9249 | 31.3085 | 31.5547 |
| `latitude` | 4.0928 | 9.8874 | 7.1659 | 7.457 |
| `value` | 1.0 | 1120000.0 | 24287.4641 | 3920.0 |

---

## 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。所有列名统一转换为小写并标准化为蛇形命名法（snake_case）。将常见的缺失值标记（`N/A`、`null`、`none`、`-`、`unknown`、`no data`、`#N/A`）统一替换为标准缺失值标记`NaN`。基于解析成功率（阈值>85%），将1列从字符串类型转换为数值型或日期时间型。使用固定随机种子（42）将数据集按80/20的比例划分为训练集与测试集，并保存为Snappy压缩的Parquet格式文件。

---

## 数据集局限性
- 本数据集源自FEWS NET，未经过非洲电羊（ESA）的独立验证。
- 自动化数据清洗流程无法修正原始数据收集中的错报值、定义不一致或抽样偏差问题。
- 以下列存在超过20%的缺失值，在建模过程中需谨慎使用：`value`。
- 请参阅[原始HDX数据集页面](https://data.humdata.org/dataset/fewsnet_staple_food_price_data_for_south_sudan_weekly_6857)查看发布机构提供的方法说明与相关注意事项。

---

## 引用格式
bibtex
@dataset{hdx_africa_fewsnet_staple_food_price_data_for_south_sudan_weekly_6857,
title = {南苏丹每周FEWS NET主粮价格数据集},
author = {FEWS NET},
year = {2026},
url = {https://data.humdata.org/dataset/fewsnet_staple_food_price_data_for_south_sudan_weekly_6857},
note = {由非洲电羊（Electric Sheep Africa）重新打包以适配机器学习需求，链接：https://huggingface.co/electricsheepafrica}
}

---

*[非洲电羊（Electric Sheep Africa）—— 非洲机器学习数据集基础设施，尼日利亚拉各斯。]*