Class IndexedDBVectorStore

IndexedDBVectorStore provides persistent vector storage using IndexedDB. Supports similarity search, MMR, and batch operations.

Index

Constructors

constructor

Methods

addDocuments addVectors clear close count maxMarginalRelevanceSearch similaritySearch similaritySearchVectorWithScore similaritySearchWithScore fromDocuments fromExistingIndex fromTexts

Constructors

constructor

new IndexedDBVectorStore(
    embeddings: EmbeddingsInterface<
        Float32Array<ArrayBufferLike>
        | number[],
    >,
    args?: IndexedDBVectorStoreArgs,
): IndexedDBVectorStore
Create an IndexedDBVectorStore instance.
Parameters
- embeddings: EmbeddingsInterface<Float32Array<ArrayBufferLike> | number[]>
  Embeddings interface for vectorization.
- args: IndexedDBVectorStoreArgs = {}
  Optional configuration for DB, store, and similarity.
Returns IndexedDBVectorStore
- Defined in store/IndexedDBVectorStore.ts:63

Methods

addDocuments

addDocuments(documents: Document<Record<string, any>>[]): Promise<void>
Add an array of documents to the store.
Parameters
- documents: Document<Record<string, any>>[]
  Documents to add.
Returns Promise<void>
- Defined in store/IndexedDBVectorStore.ts:106

addVectors

addVectors(
vectors: number[][],
documents: Document<Record<string, any>>[],
): Promise<void>
Add precomputed vectors and their documents to the store.
Parameters
- vectors: number[][]
  Array of vectors.
- documents: Document<Record<string, any>>[]
  Corresponding documents.
Returns Promise<void>
- Defined in store/IndexedDBVectorStore.ts:115

clear

clear(): Promise<void>
Clear all records from the store.

Returns Promise<void>
- Defined in store/IndexedDBVectorStore.ts:213

close

close(): Promise<void>
Close the underlying database connection.

Returns Promise<void>
- Defined in store/IndexedDBVectorStore.ts:228

count

count(): Promise<number>
Get the number of records in the store.

Returns Promise<number>
Number of records.
- Defined in store/IndexedDBVectorStore.ts:221

maxMarginalRelevanceSearch

maxMarginalRelevanceSearch(
    query: string,
    options: {
        fetchK?: number;
        filter?: IndexedDBVectorStoreFilter;
        k: number;
        lambda?: number;
    },
): Promise<Document<Record<string, any>>[]>
Perform a Maximal Marginal Relevance (MMR) search for diverse results.
Parameters
- query: string
  Query string.
- options: {
      fetchK?: number;
      filter?: IndexedDBVectorStoreFilter;
      k: number;
      lambda?: number;
  }
  MMR options (k, fetchK, lambda, filter).
Returns Promise<Document<Record<string, any>>[]>
Array of matching documents.
- Defined in store/IndexedDBVectorStore.ts:178

similaritySearch

similaritySearch(
    query: string,
    k: number,
    filter?: IndexedDBVectorStoreFilter,
): Promise<Document<Record<string, any>>[]>
Perform a similarity search for documents matching a query string.
Parameters
- query: string
  Query string.
- k: number
  Number of top results.
- Optionalfilter: IndexedDBVectorStoreFilter
  Optional filter function.
Returns Promise<Document<Record<string, any>>[]>
Array of matching documents.
- Defined in store/IndexedDBVectorStore.ts:130

similaritySearchVectorWithScore

similaritySearchVectorWithScore(
    query: ArrayLike<number>,
    k: number,
    filter?: IndexedDBVectorStoreFilter,
): Promise<[Document<Record<string, any>>, number][]>
Parameters
- query: ArrayLike<number>
- k: number
- Optionalfilter: IndexedDBVectorStoreFilter
Returns Promise<[Document<Record<string, any>>, number][]>
- Defined in store/IndexedDBVectorStore.ts:160

similaritySearchWithScore

similaritySearchWithScore(
    query: string,
    k: number,
    filter?: IndexedDBVectorStoreFilter,
): Promise<[Document<Record<string, any>>, number][]>
Perform a similarity search and return documents with their similarity scores.
Parameters
- query: string
  Query string.
- k: number
  Number of top results.
- Optionalfilter: IndexedDBVectorStoreFilter
  Optional filter function.
Returns Promise<[Document<Record<string, any>>, number][]>
Array of [Document, score] pairs.
- Defined in store/IndexedDBVectorStore.ts:151

`Static`fromDocuments

fromDocuments(
    documents: Document<Record<string, any>>[],
    embeddings: EmbeddingsInterface<Float32Array<ArrayBufferLike> | number[]>,
    args?: IndexedDBVectorStoreArgs,
): Promise<IndexedDBVectorStore>
Parameters
- documents: Document<Record<string, any>>[]
- embeddings: EmbeddingsInterface<Float32Array<ArrayBufferLike> | number[]>
- Optionalargs: IndexedDBVectorStoreArgs
Returns Promise<IndexedDBVectorStore>
- Defined in store/IndexedDBVectorStore.ts:254

`Static`fromExistingIndex

fromExistingIndex(
    embeddings: EmbeddingsInterface<
        Float32Array<ArrayBufferLike>
        | number[],
    >,
    args?: IndexedDBVectorStoreArgs,
): Promise<IndexedDBVectorStore>
Parameters
- embeddings: EmbeddingsInterface<Float32Array<ArrayBufferLike> | number[]>
- Optionalargs: IndexedDBVectorStoreArgs
Returns Promise<IndexedDBVectorStore>
- Defined in store/IndexedDBVectorStore.ts:264

`Static`fromTexts

fromTexts(
    texts: string[],
    metadatas: object | object[],
    embeddings: EmbeddingsInterface<Float32Array<ArrayBufferLike> | number[]>,
    args?: IndexedDBVectorStoreArgs,
): Promise<IndexedDBVectorStore>
Parameters
- texts: string[]
- metadatas: object | object[]
- embeddings: EmbeddingsInterface<Float32Array<ArrayBufferLike> | number[]>
- Optionalargs: IndexedDBVectorStoreArgs
Returns Promise<IndexedDBVectorStore>
- Defined in store/IndexedDBVectorStore.ts:232

Class IndexedDBVectorStore

Index

Constructors

Methods

Constructors

constructor

Parameters

Returns IndexedDBVectorStore

Methods

addDocuments

Parameters

Returns Promise<void>

addVectors

Parameters

Returns Promise<void>

clear

Returns Promise<void>

close

Returns Promise<void>

count

Returns Promise<number>

maxMarginalRelevanceSearch

Parameters

Returns Promise<Document<Record<string, any>>[]>

similaritySearch

Parameters

Returns Promise<Document<Record<string, any>>[]>

similaritySearchVectorWithScore

Parameters

Returns Promise<[Document<Record<string, any>>, number][]>

similaritySearchWithScore

Parameters

Returns Promise<[Document<Record<string, any>>, number][]>

StaticfromDocuments

Parameters

Returns Promise<IndexedDBVectorStore>

StaticfromExistingIndex

Parameters

Returns Promise<IndexedDBVectorStore>

StaticfromTexts

Parameters

Returns Promise<IndexedDBVectorStore>

Settings

On This Page

`Static`fromDocuments

`Static`fromExistingIndex

`Static`fromTexts