Supporting Microsoft’s GraphRAG: Part 3 - Graph Analysis
In the previous section, we have created a graph in TigerGraph, and loaded the CSV data into it.
Now, let’s use Jupyter Notebook to explore the graph data and perform graph analysis.
To run this Jupyter Notebook, you can download the original .ipynb file from msft_graphrag_3.ipynb.
Retrieving the Graph from TigerGraph
Since the graph has already been created in TigerGraph, redefining its schema is unnecessary. Instead, you can provide the graph name to retrieve it. TigerGraphX will verify if the graph exists in TigerGraph and, if it does, will return the corresponding graph.
Define the TigerGraph Connection Configuration
Before retrieving the graph schema, you need to configure the TigerGraph connection settings.
The recommended approach is to use environment variables, such as setting them with the export command in the shell. Here, to illustrate the demo, we configure them within Python using the os.environ method. You can find more methods for configuring connection settings in Graph.__init__.
>>> import os
>>> os.environ["TG_HOST"] = "http://127.0.0.1"
>>> os.environ["TG_USERNAME"] = "tigergraph"
>>> os.environ["TG_PASSWORD"] = "tigergraph"
Retrieve a Graph and Print Its Schema
Once the graph has been created in TigerGraph, you can retrieve it without manually defining the schema using the Graph.from_db method, which requires only the graph name:
>>> from tigergraphx import Graph
>>> G = Graph.from_db("GraphRAG")
Display the Graph Schema
Now, let's print the schema of the graph in a well-formatted manner:
>>> import json
>>> schema = G.get_schema()
>>> print(json.dumps(schema, indent=4, default=str))
{
"graph_name": "GraphRAG",
"nodes": {
"Document": {
"primary_key": "id",
"attributes": {
"id": {
"data_type": "DataType.STRING",
"default_value": null
},
"title": {
"data_type": "DataType.STRING",
"default_value": null
}
},
"vector_attributes": {}
},
"TextUnit": {
"primary_key": "id",
"attributes": {
"id": {
"data_type": "DataType.STRING",
"default_value": null
},
"text": {
"data_type": "DataType.STRING",
"default_value": null
},
"n_tokens": {
"data_type": "DataType.UINT",
"default_value": null
}
},
"vector_attributes": {}
},
"Entity": {
"primary_key": "id",
"attributes": {
"id": {
"data_type": "DataType.STRING",
"default_value": null
},
"human_readable_id": {
"data_type": "DataType.UINT",
"default_value": null
},
"name": {
"data_type": "DataType.STRING",
"default_value": null
},
"entity_type": {
"data_type": "DataType.STRING",
"default_value": null
},
"description": {
"data_type": "DataType.STRING",
"default_value": null
}
},
"vector_attributes": {
"emb_description": {
"dimension": 1536,
"index_type": "HNSW",
"data_type": "FLOAT",
"metric": "COSINE"
}
}
},
"Relationship": {
"primary_key": "id",
"attributes": {
"id": {
"data_type": "DataType.STRING",
"default_value": null
},
"human_readable_id": {
"data_type": "DataType.UINT",
"default_value": null
},
"rank": {
"data_type": "DataType.UINT",
"default_value": null
},
"weight": {
"data_type": "DataType.DOUBLE",
"default_value": null
},
"description": {
"data_type": "DataType.STRING",
"default_value": null
}
},
"vector_attributes": {}
},
"Community": {
"primary_key": "id",
"attributes": {
"id": {
"data_type": "DataType.STRING",
"default_value": null
},
"level": {
"data_type": "DataType.UINT",
"default_value": null
},
"rank": {
"data_type": "DataType.DOUBLE",
"default_value": null
},
"rank_explanation": {
"data_type": "DataType.STRING",
"default_value": null
},
"title": {
"data_type": "DataType.STRING",
"default_value": null
},
"full_content": {
"data_type": "DataType.STRING",
"default_value": null
},
"summary": {
"data_type": "DataType.STRING",
"default_value": null
}
},
"vector_attributes": {}
}
},
"edges": {
"document_contains_text_unit": {
"is_directed_edge": false,
"from_node_type": "Document",
"to_node_type": "TextUnit",
"discriminator": "set()",
"attributes": {}
},
"text_unit_contains_entity": {
"is_directed_edge": false,
"from_node_type": "TextUnit",
"to_node_type": "Entity",
"discriminator": "set()",
"attributes": {}
},
"text_unit_contains_relationship": {
"is_directed_edge": false,
"from_node_type": "TextUnit",
"to_node_type": "Relationship",
"discriminator": "set()",
"attributes": {}
},
"relationship_source": {
"is_directed_edge": false,
"from_node_type": "Relationship",
"to_node_type": "Entity",
"discriminator": "set()",
"attributes": {}
},
"relationship_target": {
"is_directed_edge": false,
"from_node_type": "Relationship",
"to_node_type": "Entity",
"discriminator": "set()",
"attributes": {}
},
"community_contains_entity": {
"is_directed_edge": false,
"from_node_type": "Community",
"to_node_type": "Entity",
"discriminator": "set()",
"attributes": {}
},
"community_contains_relationship": {
"is_directed_edge": false,
"from_node_type": "Community",
"to_node_type": "Relationship",
"discriminator": "set()",
"attributes": {}
}
}
}
Display Node and Edge Counts
Gain deeper insights into the graph by exploring details such as the total number of nodes and the count of nodes for each node type.
Display the Total Number of Nodes
>>> G.number_of_nodes()
2883
Display the Count of Nodes for Each Node Type
>>> for node_type in schema["nodes"]:
... print(f"{node_type}: {G.number_of_nodes(node_type)}")
Document: 1
TextUnit: 104
Entity: 1577
Relationship: 1092
Community: 109
Display the Total Number of Edges
>>> G.number_of_edges()
10313
Display the Count of Edges for Each Edge Type
>>> for edge_type in schema["edges"]:
... print(f"{edge_type}: {G.number_of_edges(edge_type)}")
document_contains_text_unit: 104
text_unit_contains_entity: 2095
text_unit_contains_relationship: 1282
relationship_source: 1092
relationship_target: 1092
community_contains_entity: 1956
community_contains_relationship: 2692
Retrieving Sample Nodes
Retrieve Sample Entity Nodes.
>>> print(G.get_nodes(node_type="Entity", limit=2))
v_id v_type human_readable_id entity_type \
0 c0803923646246c5a203810faa4e4464 Entity 825 GEO
1 6069e8895f924b659534f74d6736e69d Entity 830 GEO
name description \
0 VALLEY STREAM Valley Stream is a location in New York where ...
1 CHINA China is a country in East Asia where Walmart ...
id
0 c0803923646246c5a203810faa4e4464
1 6069e8895f924b659534f74d6736e69d
Retrieve Sample Relationship Nodes
>>> print(G.get_nodes(node_type="Relationship", limit=2))
v_id v_type human_readable_id rank \
0 5e7864d8153f4aa8936b253792f0b636 Relationship 1066 32
1 1db19aed7ed54b44b4e8f71b7588e0dd Relationship 1058 16
weight description \
0 8 Animax is a channel owned by Sony Pictures Tel...
1 7 Guerrilla Cambridge developed games for the Pl...
id
0 5e7864d8153f4aa8936b253792f0b636
1 1db19aed7ed54b44b4e8f71b7588e0dd
Retrieve Sample Community Nodes
>>> print(G.get_nodes(node_type="Community", limit=2))
v_id v_type summary level \
0 49 Community This community encompasses various command-lin... 1
1 61 Community The community centers around the Battles of Ma... 1
full_content rank id \
0 # DOS and Command-Line Operating Systems Commu... 7.5 49
1 # Battles of Manassas and Civil War Historians... 6.5 61
rank_explanation title
0 The impact severity rating is high due to the ... Community 49
1 The impact severity rating is moderate to high... Community 61
What’s Next?
- Supporting Microsoft’s GraphRAG: Part 4: Perform queries using GSQL and Python-native TigerGraphX, with global and local context builders.
Start transforming your GraphRAG workflows with the power of TigerGraphX today!