---
license: mit
source_datasets:
- approximatelabs/tablib-v1-sample
task_categories:
- tabular-to-graph
- graph-learning
- network-analysis
language:
- en
configs:
- config_name: default
  data_files:
  - split: part_1
    path: data/part_1-*
dataset_info:
  features:
  - name: batch_id
    dtype: int64
  - name: table_id
    dtype: int64
  - name: reconstruction_id
    dtype: int64
  - name: graph_data
    dtype: string
  - name: column_names
    dtype: string
  - name: num_nodes
    dtype: int64
  - name: num_edges
    dtype: int64
  - name: serialization
    dtype: string
  - name: original_size
    dtype: int64
  - name: serialized_size
    dtype: int64
  - name: key
    dtype: string
  splits:
  - name: part_1
    num_bytes: 90713
    num_examples: 3
  download_size: 28440
  dataset_size: 90713
---

# GraphTab Sample Dataset

This dataset contains tables from `approximatelabs/tablib-v1-sample` processed into graph representations.


## Usage

```python
from datasets import load_dataset
import graphtab

# Load the dataset
dataset = load_dataset("{full_repo_id}")

# Access a graph
graph_data = dataset['test'][0]

# Deserialize it
graph = graphtab.deserialize_graph(graph_data['graph_data'], graph_data['serialization'])
```

## Citation

If you use this dataset, please cite both the original dataset and GraphTab.