### Navigate to Evaluation Directory

Source: https://github.com/lhrlab/hypergraphrag/blob/main/evaluation/README.md

Changes the current working directory to 'evaluation' to begin the setup process for HyperGraphRAG evaluation.

```bash
cd evaluation
```

--------------------------------

### Setup HyperGraphRAG Environment

Source: https://github.com/lhrlab/hypergraphrag/blob/main/README.md

This snippet details the steps to set up a Python environment for HyperGraphRAG using Conda and install the necessary dependencies from a requirements file.

```bash
conda create -n hypergraphrag python=3.11
conda activate hypergraphrag
pip install -r requirements.txt
```

--------------------------------

### Construct Knowledge Hypergraph

Source: https://github.com/lhrlab/hypergraphrag/blob/main/evaluation/README.md

Initiates the process of building the knowledge hypergraph using the 'script_insert.py' script. It runs in the background and logs output to a file. The example shows construction for 'hypertension', with commented-out lines for other datasets.

```bash
nohup python script_insert.py --cls hypertension > result_hypertension_insert.log 2>&1 &
# nohup python script_insert.py --cls agriculture > result_agriculture_insert.log 2>&1 &
# nohup python script_insert.py --cls cs > result_cs_insert.log 2>&1 &
# nohup python script_insert.py --cls legal > result_legal_insert.log 2>&1 &
# nohup python script_insert.py --cls mix > result_mix_insert.log 2>&1 &
```

--------------------------------

### View Evaluation Results

Source: https://github.com/lhrlab/hypergraphrag/blob/main/evaluation/README.md

Displays the evaluation results using 'see_score.py', specifying the data source and the method. This command is used to review the performance metrics obtained from the previous step. Examples for different methods are commented out.

```bash
python see_score.py --data_source hypertension --method HyperGraphRAG
# python see_score.py --data_source hypertension --method StandardRAG
# python see_score.py --data_source hypertension --method NaiveGeneration
```

--------------------------------

### Evaluate Generation Performance

Source: https://github.com/lhrlab/hypergraphrag/blob/main/evaluation/README.md

Evaluates the generated content using 'get_score.py', specifying the data source and the method used for generation. It also sets the CUDA visible devices. The example evaluates 'HyperGraphRAG' for 'hypertension'.

```bash
CUDA_VISIBLE_DEVICES=0 python get_score.py --data_source hypertension --method HyperGraphRAG
# CUDA_VISIBLE_DEVICES=0 python get_score.py --data_source hypertension --method StandardRAG
# CUDA_VISIBLE_DEVICES=0 python get_score.py --data_source hypertension --method NaiveGeneration
```

--------------------------------

### Generate Based on Retrieved Knowledge

Source: https://github.com/lhrlab/hypergraphrag/blob/main/evaluation/README.md

Generates content based on retrieved knowledge using 'get_generation.py'. It specifies the data sources and the generation method. The example shows generation for 'hypertension' using 'HyperGraphRAG', with other methods commented out.

```bash
python get_generation.py --data_sources hypertension --methods HyperGraphRAG
# python get_generation.py --data_sources hypertension --methods StandardRAG
# python get_generation.py --data_sources hypertension --methods NaiveGeneration
```

--------------------------------

### Retrieve Knowledge with HyperGraphRAG

Source: https://github.com/lhrlab/hypergraphrag/blob/main/evaluation/README.md

Retrieves knowledge using the 'script_hypergraphrag.py' script, specifying the data source. This command is part of the retrieval step in the evaluation process. Alternative retrieval methods are commented out.

```bash
python script_hypergraphrag.py --data_source hypertension
# python script_standardrag.py --data_source hypertension
# python script_naivegeneration.py --data_source hypertension
```

--------------------------------

### Construct Knowledge HyperGraph with HyperGraphRAG

Source: https://github.com/lhrlab/hypergraphrag/blob/main/README.md

Demonstrates how to initialize HyperGraphRAG and insert unique contexts to build a knowledge hypergraph. It requires setting the OpenAI API key and providing context data in a JSON file.

```python
import os
import json
from hypergraphrag import HyperGraphRAG
os.environ["OPENAI_API_KEY"] = "your_openai_api_key"

rag = HyperGraphRAG(working_dir=f"expr/example")

with open(f"example_contexts.json", mode="r") as f:
    unique_contexts = json.load(f)
    
rag.insert(unique_contexts)
```

--------------------------------

### Query Knowledge HyperGraph with HyperGraphRAG

Source: https://github.com/lhrlab/hypergraphrag/blob/main/README.md

Shows how to query the constructed knowledge hypergraph using HyperGraphRAG. This involves initializing the RAG system with the working directory and providing a natural language query.

```python
import os
from hypergraphrag import HyperGraphRAG
os.environ["OPENAI_API_KEY"] = "your_openai_api_key"

rag = HyperGraphRAG(working_dir=f"expr/example")

query_text = 'How strong is the evidence supporting a systolic BP target of 120–129 mmHg in elderly or frail patients, considering potential risks like orthostatic hypotension, the balance between cardiovascular benefits and adverse effects, and the feasibility of implementation in diverse healthcare settings?'

result = rag.query(query_text)
print(result)
```

--------------------------------

### HyperGraphRAG Citation

Source: https://github.com/lhrlab/hypergraphrag/blob/main/README.md

Provides the BibTeX entry for citing the HyperGraphRAG paper, which details the Retrieval-Augmented Generation via Hypergraph-Structured Knowledge Representation.

```bibtex
@misc{luo2025hypergraphrag,
      title={HyperGraphRAG: Retrieval-Augmented Generation via Hypergraph-Structured Knowledge Representation},
      author={Haoran Luo and Haihong E and Guanting Chen and Yandan Zheng and Xiaobao Wu and Yikai Guo and Qika Lin and Yu Feng and Zemin Kuang and Meina Song and Yifan Zhu and Luu Anh Tuan},
      year={2025},
      eprint={2503.21322},
      archivePrefix={arXiv},
      primaryClass={cs.AI},
      url={https://arxiv.org/abs/2503.21322},
}
```

=== COMPLETE CONTENT === This response contains all available snippets from this library. No additional content exists. Do not make further requests.