Overview

Vector creation operations enable you to store and manage vector embeddings within your graph database. These operations are essential for implementing advanced AI features such as semantic search, similarity matching, recommendation systems, and machine learning applications. Understanding vector creation is crucial for building intelligent applications that can process and analyze high-dimensional data effectively.

Key capabilities include:

Creating individual vectors with custom embeddings and metadata
Performing bulk vector creation for efficient batch processing
Configuring vector models and dimensionality settings
Associating vectors with nodes or edges in your graph
Managing vector content and metadata for search optimization

These operations support various use cases such as semantic search, recommendation engines, content similarity analysis, and AI-powered data processing applications.

Create Single Vector

Create a single vector using PUT: /v1.0/tenants/{tenant-guid}/vectors. This endpoint allows you to store a vector embedding with associated metadata, model information, and content for advanced search and analysis capabilities.

curl --location --request PUT 'http://localhost:8701/v1.0/tenants/00000000-0000-0000-0000-000000000000/vectors' \
--header 'content-type: application/json' \
--header 'Authorization: ••••••' \
--data '{
    "GraphGUID": "00000000-0000-0000-0000-000000000000",
    "NodeGUID": null,
    "EdgeGUID": null,
    "Model": "all-MiniLM-L6-v2",
    "Dimensionality": 384,
    "Content": "test",
    "Vectors": [
        0.1,
        0.2,
        0.3
    ]
}'

import { LiteGraphSdk } from "litegraphdb";

var api = new LiteGraphSdk(
  "http://localhost:8701/",
  "<Tenant-Guid>",
  "*******"
);

const createVector = async () => {
  try {
    const data = await api.Vector.create({
      GraphGUID: "<graph-guid>",
      NodeGUID: "<node-guid>",
      EdgeGUID: "<edge-guid>",
      Model: "all-MiniLM-L6-v2",
      Dimensionality: 384,
      Content: "test",
      Vectors: [0.1, 0.2, 0.3],
    });
    console.log(data, "check data");
  } catch (err) {
    console.log("err:", JSON.stringify(err));
  }
};

import litegraph

sdk = litegraph.configure(
    endpoint="http://localhost:8701",
    tenant_guid="Tenant-Guid",
    access_key="******",
)

def create_vector():
    vector = litegraph.Vector.create(
        vectors=[0.1, 0.2, 0.3],
        content="Test Content",
        graph_guid="00000000-0000-0000-0000-000000000000",
        dimensionality=3,
        model="all-MiniLM-L6-v2"
    )
    print(vector)

create_vector()

using LiteGraph;
using LiteGraph.GraphRepositories.Sqlite;

LiteGraphClient liteGraph = new LiteGraphClient(new SqliteGraphRepository("litegraph.db"));
liteGraph.InitializeRepository();
VectorMetadata response = liteGraph.Vector.Create(new VectorMetadata()
{
    GraphGUID = Guid.Parse("<graph-guid>"),
    NodeGUID = Guid.Parse("<node-guid>"),
    EdgeGUID = Guid.Parse("<edge-guid>"),
    Model = "all-MiniLM-L6-v2",
    Dimensionality= 384,
    Content= "test",
    Vectors= new List<float>() { 0.1f, 0.2f, 0.3f  }
});

Response

Upon successful vector creation, the API returns a 201 Created status code with the created vector object in the response body.

{
  "GUID": "00000000-0000-0000-0000-000000000000",
  "TenantGUID": "00000000-0000-0000-0000-000000000000",
  "GraphGUID": "00000000-0000-0000-0000-000000000000",
  "NodeGUID": null,
  "EdgeGUID": null,
  "Model": "all-MiniLM-L6-v2",
  "Dimensionality": 384,
  "Content": "test",
  "Vectors": [0.1, 0.2, 0.3],
  "CreatedUtc": "2024-12-27T18:12:38.653402Z",
  "LastUpdateUtc": "2024-12-27T18:12:38.653402Z"
}

Create Multiple Vectors

Create multiple vectors in a single operation using PUT: /v1.0/tenants/{tenant-guid}/vectors/bulk. This bulk operation allows you to efficiently store multiple vector embeddings with different configurations in one API call, which is useful for batch processing and large-scale vector ingestion.

curl --location --request PUT 'http://localhost:8701/v1.0/tenants/00000000-0000-0000-0000-000000000000/vectors/bulk' \
--header 'content-type: application/json' \
--header 'Authorization: ••••••' \
--data '[
  {
    "GraphGUID": "00000000-0000-0000-0000-000000000000",
    "NodeGUID": null,
    "EdgeGUID": null,
    "Model": "all-MiniLM-L6-v2",
    "Dimensionality": 384,
    "Content": "test",
    "Vectors": [
      0.1,
      0.2,
      0.3
    ]
  }
]'

import { LiteGraphSdk } from "litegraphdb";

var api = new LiteGraphSdk(
  "http://localhost:8701/",
  "<Tenant-Guid>",
  "*******"
);

const createMultipleVectors = async () => {
  try {
    const data = await api.Vector.createBulk([
      {
        GraphGUID: "<graph-guid>",
        NodeGUID: "<node-guid>",
        EdgeGUID: "<edge-guid>",
        Model: "all-MiniLM-L6-v2",
        Dimensionality: 384,
        Content: "test 1",
        Vectors: [0.1, 0.2, 0.3],
      },
      {
        GraphGUID: "<graph-guid>",
        NodeGUID: "<node-guid>",
        EdgeGUID: "<edge-guid>",
        Model: "all-MiniLM-L6-v2",
        Dimensionality: 390,
        Content: "test 2",
        Vectors: [0.5, 0.7, 0.9],
      },
    ]);
    console.log(data, "check data");
  } catch (err) {
    console.log("err:", JSON.stringify(err));
  }
};

import litegraph

sdk = litegraph.configure(
    endpoint="http://localhost:8701",
    tenant_guid="Tenant-Guid",
    access_key="******",
)

def create_multiple_vector():
    vectors = litegraph.Vector.create_multiple([
        {
            "graph_guid": "00000000-0000-0000-0000-000000000000",
            "node_guid": None,
            "edge_guid": None,
            "model": "all-MiniLM-L6-v2",
            "dimensionality": 384,
            "content": "Test Content",
            "vectors": [0.1, 0.2, 0.3]
        },
        {
            "graph_guid": "00000000-0000-0000-0000-000000000000",
            "node_guid": None,
            "edge_guid": None,
            "model": "all-MiniLM-L6-v2",
            "dimensionality": 384,
            "content": "Test Content 2",
            "vectors": [0.4, 0.5, 0.6]
        }
    ])
    print(vectors)

create_multiple_vector()

using LiteGraph;
using LiteGraph.GraphRepositories.Sqlite;

LiteGraphClient liteGraph = new LiteGraphClient(new SqliteGraphRepository("litegraph.db"));
liteGraph.InitializeRepository();
List<VectorMetadata> response = liteGraph.Vector.CreateMany(Guid.Parse("<tenant-guid>"), new List<VectorMetadata>()
{ 
    new VectorMetadata()
    {
        GraphGUID = Guid.Parse("<graph-guid>"),
        NodeGUID = Guid.Parse("<node-guid>"),
        EdgeGUID = Guid.Parse("<edge-guid>"),
        Model = "all-MiniLM-L6-v2",
        Dimensionality= 384,
        Content= "test",
        Vectors= new List<float>() { 0.1f, 0.2f, 0.3f }
    }
});

Response

Upon successful bulk vector creation, the API returns a 201 Created status code with an array of created vector objects in the response body.

[
  {
    "GUID": "00000000-0000-0000-0000-000000000001",
    "TenantGUID": "00000000-0000-0000-0000-000000000000",
    "GraphGUID": "00000000-0000-0000-0000-000000000000",
    "NodeGUID": null,
    "EdgeGUID": null,
    "Model": "all-MiniLM-L6-v2",
    "Dimensionality": 384,
    "Content": "test",
    "Vectors": [0.1, 0.2, 0.3],
    "CreatedUtc": "2024-12-27T18:12:38.653402Z",
    "LastUpdateUtc": "2024-12-27T18:12:38.653402Z"
  }
]

Best Practices

When creating vectors, consider the following recommendations:

Choose Appropriate Models: Select vector models that match your use case and data type
Validate Dimensionality: Ensure vector dimensions match your model's expected output
Optimize Content: Use meaningful content descriptions for better search results
Batch Operations: Use bulk creation for multiple vectors to improve performance

Next Steps

After successfully creating vectors, consider these next actions:

Configure Vector Indexes: Set up HNSW indexes for efficient similarity search
Perform Vector Search: Use the created vectors for semantic search operations
Update Vectors: Modify existing vectors as your data evolves
Monitor Performance: Track vector search performance and optimize as needed
Integrate with Applications: Use vectors in your AI-powered applications
Backup Vector Data: Implement backup strategies for critical vector data