---
license: mit
task_categories:
  - tabular-regression
  - time-series-forecasting
language:
  - en
tags:
  - real-estate
  - housing
  - australia
  - nsw
  - property-prices
size_categories:
  - 1M<n<10M
---

# Housing NSW Dataset

## Dataset Description

Comprehensive property market data for New South Wales, Australia.

### Contents

- **Property Sales** (`nsw_property_sales_master.csv`): 3.49M historical property transactions from 1990-2024
- **Enriched Properties** (`nsw_property_master_enriched.csv`): 83.5K properties with detailed analytics including market indicators, demographic data, and transport accessibility  
- **Rental Market** (`nsw_rental_market_master.csv`): 48.2K rental bond records with pricing data

### Key Features

- Property prices and sale dates
- Property characteristics (bedrooms, bathrooms, land area)
- Location data (suburb, postcode, some with lat/lon)
- Market analytics (median prices, growth potential, rental yields)
- Demographic indicators (income, population)
- Transport accessibility scores
- School ratings and crime indices

## Usage

```python
from datasets import load_dataset

# Load the dataset
dataset = load_dataset("thehooklab/housing-nsw")

# Access the data
sales_data = dataset['train']  # All data is in the 'train' split
```

## Data Quality

- All numeric columns have been validated and non-numeric values converted to NaN
- String columns are properly encoded
- Files are tested for PyArrow compatibility

## License

MIT License

## Citation

If you use this dataset, please cite:

```
@dataset{housing_nsw_2024,
  title={Housing NSW Dataset},
  author={TheHookLab},
  year={2024},
  publisher={HuggingFace}
}
```