---
pretty_name: ArchEGraph-demo
license: apache-2.0
task_categories:
- graph-ml
- time-series-forecasting
language:
- en
tags:
- building-energy
- simulation
- graph
- weather
- demo
size_categories:
- n<1K
configs:
- config_name: manifest
  default: true
  data_files:
  - split: train
    path: manifest.csv
- config_name: split_demo
  data_files:
  - split: train
    path: split/split_demo.csv
- config_name: split_demo_mesh
  data_files:
  - split: train
    path: split/split_demo_mesh.csv
---

# ArchEGraph-demo

ArchEGraph-demo is a compact demo package of the ArchEGraph building-energy dataset for graph-based and weather-conditioned learning.

## Dataset Summary

- Total cases in `manifest.csv`: 300
- Unique buildings: 75
- Unique weather IDs: 48
- `n_steps`: always 8,760
- `n_spaces` range: 2 to 132

This package currently stores:

- `manifest.csv` (index of all demo cases)
- `building/` (75 files)
- `geometry/` (75 files)
- `weather/` (48 files)
- `energy/` (300 files)
- `split/` (demo split CSV files)

## Data Layout

Each row in `manifest.csv` contains:

- `sample_id`: case ID (`building__weather` style)
- `source_job_tag`: source identifier
- `weather_id`: weather/location key
- `building_id`: building key
- `energy_file`: relative path to energy npz file under `energy/`
- `n_steps`: number of time steps
- `n_spaces`: number of spaces/zones

## Included Split Files

- `split/split_demo.csv` (300 rows)
- `split/split_demo_mesh.csv` (300 rows)

`split/split_demo.csv` uses these columns:

- `case_id`, `sample_id`, `building_id`, `weather_id`, `subset`, `split`, `scenario`

`split/split_demo_mesh.csv` uses these columns:

- `building_id`, `split`

## Quick Start

```python
import pandas as pd
from pathlib import Path

root = Path(".")  # dataset root
manifest = pd.read_csv(root / "manifest.csv")

row = manifest.iloc[0]
energy_path = root / "energy" / row["energy_file"]
building_path = root / "building" / f"{row['building_id']}.npz"
weather_path = root / "weather" / f"{row['weather_id']}.npz"

print(row["sample_id"])
print(energy_path, building_path, weather_path)
```

## Notes

- This repository is the demo package, not the full PACK release.
- Energy files in this demo package are referenced by `energy_file` from `manifest.csv`.
- `split/split_demo.csv` and `split/split_demo_mesh.csv` provide ready-to-use predefined splits for the packaged demo samples.

## Citation

If you use this dataset, please cite your project/paper and this Hugging Face dataset page.