# AgriYield: Predict Crop Yield from Soil, Weather — Data

来源：https://www.kaggle.com/competitions/agriyield-2025/data

采集时间：2026-10-20T05:08:05Z

## Dataset Files You Need

You must prepare **3 required files**:

| File Name               | Description                                   |
| ----------------------- | --------------------------------------------- |
| `train.csv`             | Training data with features + target (yield)  |
| `test.csv`              | Test data with only features                  |
| `sample_submission.csv` | Format participants use to submit predictions |

---

##  Recommended Dataset Structure

```plaintext
/dataset/
│
├── train.csv
├── test.csv
├── sample_submission.csv
└── data_description.md  ← optional, but highly recommended
```

# Data Dictionary

## Features

- `field_id`: Unique identifier for each field
- `soil_ph`: Soil acidity (pH scale)
- `organic_matter`: Percentage of organic matter in the soil
- `sand_pct`: Percentage of sand in soil composition
- `temperature`: Average temperature (°C) during the growing season
- `humidity`: Average relative humidity (%) during growing season
- `rainfall`: Total rainfall (mm) during growing season
- `ndvi`: Normalized Difference Vegetation Index (remote sensing)

## Target

- `yield`: Maize yield measured in kilograms per hectare (kg/ha)