Arko007/walnut-rancidity-predictor

---
license: mit
language:
- en
pretty_name: Walnut Storage Timeseries (Indian Conditions)
size_categories:
- 1M<n<10M
task_categories:
- time-series-forecasting
tags:
- food-science
- walnut
- rancidity
- oxidation
- arrhenius
- storage
- india
- synthetic
- shelf-life
dataset_info:
features:
- name: sequence_id
dtype: int32
- name: day
dtype: int16
- name: temperature
dtype: float32
- name: humidity
dtype: float32
- name: moisture
dtype: float32
- name: oxygen
dtype: float32
- name: peroxide_value
dtype: float32
- name: free_fatty_acids
dtype: float32
- name: hexanal_level
dtype: float32
- name: oxidation_index
dtype: float32
- name: rancidity_probability
dtype: float32
- name: shelf_life_remaining_days
dtype: float32
- name: decay_curve_value
dtype: float32
splits:
- name: train
num_examples: 5392174
num_rows: 5392174
---

# Walnut Storage Timeseries Dataset (Indian Conditions)

Synthetic time-series dataset for training walnut rancidity prediction models.
Generated using **Arrhenius-based lipid oxidation kinetics** simulating real Indian storage scenarios.

## Dataset Details

| Property | Value |
|---|---|
| Total rows | 5,392,174 |
| Total sequences | 90,000 |
| Sequence length | 30–90 days |
| Storage duration | 0–180 days |
| Generation method | Arrhenius kinetics + noise |

## Storage Scenarios

| Scenario | Temperature | Humidity | Weight |
|---|---|---|---|
| Cold storage warehouse | 2–8 °C | 40–60 % | 20 % |
| Hill region (Kashmir / Himachal Pradesh) | 5–20 °C | 35–65 % | 25 % |
| Ambient warehouse | 18–32 °C | 50–75 % | 30 % |
| Hot transport (Indian summer) | 28–40 °C | 55–85 % | 25 % |

## Chemistry Model

Oxidation simulated via Arrhenius kinetics:

```
k = A · exp(-Ea / (R · T)) A=1.5e12, Ea=80000 J/mol, R=8.314
PV(t) = PV₀ · exp(k · t)
```

Rancidity threshold: **PV > 5 meq/kg**

## Columns

| Column | Unit | Description |
|---|---|---|
| `sequence_id` | — | Unique storage sequence identifier |
| `day` | days | Storage day index |
| `temperature` | °C | Daily ambient temperature |
| `humidity` | % RH | Relative humidity |
| `moisture` | % | Walnut moisture content |
| `oxygen` | fraction | Oxygen level (0.18–0.23) |
| `peroxide_value` | meq/kg | Primary oxidation marker |
| `free_fatty_acids` | % | FFA content |
| `hexanal_level` | ppm | Volatile oxidation byproduct |
| `oxidation_index` | — | Composite oxidation score |
| `rancidity_probability` | [0,1] | sigmoid(PV − 5) |
| `shelf_life_remaining_days` | days | Days until PV exceeds 5 meq/kg |
| `decay_curve_value` | [0,1] | Normalised PV (PV/10, capped at 1) |

## Associated Model

Trained model: [Arko007/walnut-rancidity-predictor](https://huggingface.co/Arko007/walnut-rancidity-predictor)

## License

MIT