1.

Current, the stable version of DHG is 0.9.4. You can install it with pip as follows: python pip install dhg pip install -r ./docs/requirements.txt

2.

• Improved hypergraph neural network model: ./HyperRCA/DHG/models/hypergraphs/hgnnp.py
• Improved hypergraph neural network convolution layer: ./HyperRCA/DHG/nn/convs/hypergraphs/hgnnp_conv.py
• Data loading class (mainly used for loading datasets): ./HyperRCA/DHG/utils/data_handler.py
• Main experiment class (mainly used for fault diagnosis and root cause localization): ./HyperRCA/HyperRCA.py Running HyperRCA.py will start the experiment.

Here's the English translation:

3.
1. ./HyperRCA/DHG/utils/data_handler.py python # Get and sort all files - new_gaia self.feature_files = sorted(glob.glob(os.path.join(data_root, "features_*.pkl"))) self.label_files = sorted(glob.glob(os.path.join(data_root, "labels_*.pkl"))) self.edge_list_files = sorted(glob.glob(os.path.join(data_root, "edge_list_30", "edge_list_*.pkl"))) self.num_vertices = 10 Or python # Get and sort all files - trainticket self.feature_files = sorted(glob.glob(os.path.join(data_root, "features_*.pkl"))) self.label_files = sorted(glob.glob(os.path.join(data_root, "labels_*.pkl"))) self.edge_list_files = sorted(glob.glob(os.path.join(data_root, "edge_list_30_noapi_test", "edge_list_*.pkl"))) self.num_vertices = 26

2. ./HyperRCA/HyperRCA.py
```python
"data_root": "../../datasets/new_gaia",
```
Or
```python
"data_root": "../../datasets/trainticket",
```

Thank you for the open source code https://github.com/iMoonLab/DeepHypergraph for its help with this experiment.