Skip to content

Elasticsearch

ElasticsearchStore #

Bases: BasePydanticVectorStore

Elasticsearch vector store.

Parameters:

Name Type Description Default
index_name str

Name of the Elasticsearch index.

required
es_client Optional[Any]

Optional. Pre-existing AsyncElasticsearch client.

None
es_url Optional[str]

Optional. Elasticsearch URL.

None
es_cloud_id Optional[str]

Optional. Elasticsearch cloud ID.

None
es_api_key Optional[str]

Optional. Elasticsearch API key.

None
es_user Optional[str]

Optional. Elasticsearch username.

None
es_password Optional[str]

Optional. Elasticsearch password.

None
text_field str

Optional. Name of the Elasticsearch field that stores the text.

'content'
vector_field str

Optional. Name of the Elasticsearch field that stores the embedding.

'embedding'
batch_size int

Optional. Batch size for bulk indexing. Defaults to 200.

200
distance_strategy Optional[DISTANCE_STRATEGIES]

Optional. Distance strategy to use for similarity search. Defaults to "COSINE".

'COSINE'

Raises:

Type Description
ConnectionError

If AsyncElasticsearch client cannot connect to Elasticsearch.

ValueError

If neither es_client nor es_url nor es_cloud_id is provided.

Examples:

pip install llama-index-vector-stores-elasticsearch

from llama_index.vector_stores import ElasticsearchStore

# Additional setup for ElasticsearchStore class
index_name = "my_index"
es_url = "http://localhost:9200"
es_cloud_id = "<cloud-id>"  # Found within the deployment page
es_user = "elastic"
es_password = "<password>"  # Provided when creating deployment or can be reset
es_api_key = "<api-key>"  # Create an API key within Kibana (Security -> API Keys)

# Connecting to ElasticsearchStore locally
es_local = ElasticsearchStore(
    index_name=index_name,
    es_url=es_url,
)

# Connecting to Elastic Cloud with username and password
es_cloud_user_pass = ElasticsearchStore(
    index_name=index_name,
    es_cloud_id=es_cloud_id,
    es_user=es_user,
    es_password=es_password,
)

# Connecting to Elastic Cloud with API Key
es_cloud_api_key = ElasticsearchStore(
    index_name=index_name,
    es_cloud_id=es_cloud_id,
    es_api_key=es_api_key,
)
Source code in llama-index-integrations/vector_stores/llama-index-vector-stores-elasticsearch/llama_index/vector_stores/elasticsearch/base.py
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
class ElasticsearchStore(BasePydanticVectorStore):
    """Elasticsearch vector store.

    Args:
        index_name: Name of the Elasticsearch index.
        es_client: Optional. Pre-existing AsyncElasticsearch client.
        es_url: Optional. Elasticsearch URL.
        es_cloud_id: Optional. Elasticsearch cloud ID.
        es_api_key: Optional. Elasticsearch API key.
        es_user: Optional. Elasticsearch username.
        es_password: Optional. Elasticsearch password.
        text_field: Optional. Name of the Elasticsearch field that stores the text.
        vector_field: Optional. Name of the Elasticsearch field that stores the
                    embedding.
        batch_size: Optional. Batch size for bulk indexing. Defaults to 200.
        distance_strategy: Optional. Distance strategy to use for similarity search.
                        Defaults to "COSINE".

    Raises:
        ConnectionError: If AsyncElasticsearch client cannot connect to Elasticsearch.
        ValueError: If neither es_client nor es_url nor es_cloud_id is provided.

    Examples:
        `pip install llama-index-vector-stores-elasticsearch`

        ```python
        from llama_index.vector_stores import ElasticsearchStore

        # Additional setup for ElasticsearchStore class
        index_name = "my_index"
        es_url = "http://localhost:9200"
        es_cloud_id = "<cloud-id>"  # Found within the deployment page
        es_user = "elastic"
        es_password = "<password>"  # Provided when creating deployment or can be reset
        es_api_key = "<api-key>"  # Create an API key within Kibana (Security -> API Keys)

        # Connecting to ElasticsearchStore locally
        es_local = ElasticsearchStore(
            index_name=index_name,
            es_url=es_url,
        )

        # Connecting to Elastic Cloud with username and password
        es_cloud_user_pass = ElasticsearchStore(
            index_name=index_name,
            es_cloud_id=es_cloud_id,
            es_user=es_user,
            es_password=es_password,
        )

        # Connecting to Elastic Cloud with API Key
        es_cloud_api_key = ElasticsearchStore(
            index_name=index_name,
            es_cloud_id=es_cloud_id,
            es_api_key=es_api_key,
        )
        ```

    """

    stores_text: bool = True
    index_name: str
    es_client: Optional[Any]
    es_url: Optional[str]
    es_cloud_id: Optional[str]
    es_api_key: Optional[str]
    es_user: Optional[str]
    es_password: Optional[str]
    text_field: str = "content"
    vector_field: str = "embedding"
    batch_size: int = 200
    distance_strategy: Optional[DISTANCE_STRATEGIES] = "COSINE"

    _client = PrivateAttr()

    def __init__(
        self,
        index_name: str,
        es_client: Optional[Any] = None,
        es_url: Optional[str] = None,
        es_cloud_id: Optional[str] = None,
        es_api_key: Optional[str] = None,
        es_user: Optional[str] = None,
        es_password: Optional[str] = None,
        text_field: str = "content",
        vector_field: str = "embedding",
        batch_size: int = 200,
        distance_strategy: Optional[DISTANCE_STRATEGIES] = "COSINE",
    ) -> None:
        nest_asyncio.apply()

        if es_client is not None:
            self._client = es_client.options(
                headers={"user-agent": self.get_user_agent()}
            )
        elif es_url is not None or es_cloud_id is not None:
            self._client = _get_elasticsearch_client(
                es_url=es_url,
                username=es_user,
                password=es_password,
                cloud_id=es_cloud_id,
                api_key=es_api_key,
            )
        else:
            raise ValueError(
                """Either provide a pre-existing AsyncElasticsearch or valid \
                credentials for creating a new connection."""
            )
        super().__init__(
            index_name=index_name,
            es_client=es_client,
            es_url=es_url,
            es_cloud_id=es_cloud_id,
            es_api_key=es_api_key,
            es_user=es_user,
            es_password=es_password,
            text_field=text_field,
            vector_field=vector_field,
            batch_size=batch_size,
            distance_strategy=distance_strategy,
        )

    @property
    def client(self) -> Any:
        """Get async elasticsearch client."""
        return self._client

    @staticmethod
    def get_user_agent() -> str:
        """Get user agent for elasticsearch client."""
        import llama_index.core

        version = getattr(llama_index.core, "__version__", "")
        return f"llama_index-py-vs/{version}"

    async def _create_index_if_not_exists(
        self, index_name: str, dims_length: Optional[int] = None
    ) -> None:
        """Create the AsyncElasticsearch index if it doesn't already exist.

        Args:
            index_name: Name of the AsyncElasticsearch index to create.
            dims_length: Length of the embedding vectors.
        """
        if await self.client.indices.exists(index=index_name):
            logger.debug(f"Index {index_name} already exists. Skipping creation.")

        else:
            if dims_length is None:
                raise ValueError(
                    "Cannot create index without specifying dims_length "
                    "when the index doesn't already exist. We infer "
                    "dims_length from the first embedding. Check that "
                    "you have provided an embedding function."
                )

            if self.distance_strategy == "COSINE":
                similarityAlgo = "cosine"
            elif self.distance_strategy == "EUCLIDEAN_DISTANCE":
                similarityAlgo = "l2_norm"
            elif self.distance_strategy == "DOT_PRODUCT":
                similarityAlgo = "dot_product"
            else:
                raise ValueError(f"Similarity {self.distance_strategy} not supported.")

            index_settings = {
                "mappings": {
                    "properties": {
                        self.vector_field: {
                            "type": "dense_vector",
                            "dims": dims_length,
                            "index": True,
                            "similarity": similarityAlgo,
                        },
                        self.text_field: {"type": "text"},
                        "metadata": {
                            "properties": {
                                "document_id": {"type": "keyword"},
                                "doc_id": {"type": "keyword"},
                                "ref_doc_id": {"type": "keyword"},
                            }
                        },
                    }
                }
            }

            logger.debug(
                f"Creating index {index_name} with mappings {index_settings['mappings']}"
            )
            await self.client.indices.create(index=index_name, **index_settings)

    def add(
        self,
        nodes: List[BaseNode],
        *,
        create_index_if_not_exists: bool = True,
        **add_kwargs: Any,
    ) -> List[str]:
        """Add nodes to Elasticsearch index.

        Args:
            nodes: List of nodes with embeddings.
            create_index_if_not_exists: Optional. Whether to create
                                        the Elasticsearch index if it
                                        doesn't already exist.
                                        Defaults to True.

        Returns:
            List of node IDs that were added to the index.

        Raises:
            ImportError: If elasticsearch['async'] python package is not installed.
            BulkIndexError: If AsyncElasticsearch async_bulk indexing fails.
        """
        return asyncio.get_event_loop().run_until_complete(
            self.async_add(nodes, create_index_if_not_exists=create_index_if_not_exists)
        )

    async def async_add(
        self,
        nodes: List[BaseNode],
        *,
        create_index_if_not_exists: bool = True,
        **add_kwargs: Any,
    ) -> List[str]:
        """Asynchronous method to add nodes to Elasticsearch index.

        Args:
            nodes: List of nodes with embeddings.
            create_index_if_not_exists: Optional. Whether to create
                                        the AsyncElasticsearch index if it
                                        doesn't already exist.
                                        Defaults to True.

        Returns:
            List of node IDs that were added to the index.

        Raises:
            ImportError: If elasticsearch python package is not installed.
            BulkIndexError: If AsyncElasticsearch async_bulk indexing fails.
        """
        if len(nodes) == 0:
            return []

        if create_index_if_not_exists:
            dims_length = len(nodes[0].get_embedding())
            await self._create_index_if_not_exists(
                index_name=self.index_name, dims_length=dims_length
            )

        embeddings: List[List[float]] = []
        texts: List[str] = []
        metadatas: List[dict] = []
        ids: List[str] = []
        for node in nodes:
            ids.append(node.node_id)
            embeddings.append(node.get_embedding())
            texts.append(node.get_content(metadata_mode=MetadataMode.NONE))
            metadatas.append(node_to_metadata_dict(node, remove_text=True))

        requests = []
        return_ids = []

        for i, text in enumerate(texts):
            metadata = metadatas[i] if metadatas else {}
            _id = ids[i] if ids else str(uuid.uuid4())
            request = {
                "_op_type": "index",
                "_index": self.index_name,
                self.vector_field: embeddings[i],
                self.text_field: text,
                "metadata": metadata,
                "_id": _id,
            }
            requests.append(request)
            return_ids.append(_id)

        async with self.client as client:
            await async_bulk(client, requests, chunk_size=self.batch_size, refresh=True)
            try:
                success, failed = await async_bulk(
                    client, requests, stats_only=True, refresh=True
                )
                logger.debug(
                    f"Added {success} and failed to add {failed} texts to index"
                )

                logger.debug(f"added texts {ids} to index")
                return return_ids
            except BulkIndexError as e:
                logger.error(f"Error adding texts: {e}")
                firstError = e.errors[0].get("index", {}).get("error", {})
                logger.error(f"First error reason: {firstError.get('reason')}")
                raise

    def delete(self, ref_doc_id: str, **delete_kwargs: Any) -> None:
        """Delete node from Elasticsearch index.

        Args:
            ref_doc_id: ID of the node to delete.
            delete_kwargs: Optional. Additional arguments to
                        pass to Elasticsearch delete_by_query.

        Raises:
            Exception: If Elasticsearch delete_by_query fails.
        """
        return asyncio.get_event_loop().run_until_complete(
            self.adelete(ref_doc_id, **delete_kwargs)
        )

    async def adelete(self, ref_doc_id: str, **delete_kwargs: Any) -> None:
        """Async delete node from Elasticsearch index.

        Args:
            ref_doc_id: ID of the node to delete.
            delete_kwargs: Optional. Additional arguments to
                        pass to AsyncElasticsearch delete_by_query.

        Raises:
            Exception: If AsyncElasticsearch delete_by_query fails.
        """
        try:
            async with self.client as client:
                res = await client.delete_by_query(
                    index=self.index_name,
                    query={"term": {"metadata.ref_doc_id": ref_doc_id}},
                    refresh=True,
                    **delete_kwargs,
                )
            if res["deleted"] == 0:
                logger.warning(f"Could not find text {ref_doc_id} to delete")
            else:
                logger.debug(f"Deleted text {ref_doc_id} from index")
        except Exception:
            logger.error(f"Error deleting text: {ref_doc_id}")
            raise

    def query(
        self,
        query: VectorStoreQuery,
        custom_query: Optional[
            Callable[[Dict, Union[VectorStoreQuery, None]], Dict]
        ] = None,
        es_filter: Optional[List[Dict]] = None,
        **kwargs: Any,
    ) -> VectorStoreQueryResult:
        """Query index for top k most similar nodes.

        Args:
            query_embedding (List[float]): query embedding
            custom_query: Optional. custom query function that takes in the es query
                        body and returns a modified query body.
                        This can be used to add additional query
                        parameters to the Elasticsearch query.
            es_filter: Optional. Elasticsearch filter to apply to the
                        query. If filter is provided in the query,
                        this filter will be ignored.

        Returns:
            VectorStoreQueryResult: Result of the query.

        Raises:
            Exception: If Elasticsearch query fails.

        """
        return asyncio.get_event_loop().run_until_complete(
            self.aquery(query, custom_query, es_filter, **kwargs)
        )

    async def aquery(
        self,
        query: VectorStoreQuery,
        custom_query: Optional[
            Callable[[Dict, Union[VectorStoreQuery, None]], Dict]
        ] = None,
        es_filter: Optional[List[Dict]] = None,
        **kwargs: Any,
    ) -> VectorStoreQueryResult:
        """Asynchronous query index for top k most similar nodes.

        Args:
            query_embedding (VectorStoreQuery): query embedding
            custom_query: Optional. custom query function that takes in the es query
                        body and returns a modified query body.
                        This can be used to add additional query
                        parameters to the AsyncElasticsearch query.
            es_filter: Optional. AsyncElasticsearch filter to apply to the
                        query. If filter is provided in the query,
                        this filter will be ignored.

        Returns:
            VectorStoreQueryResult: Result of the query.

        Raises:
            Exception: If AsyncElasticsearch query fails.

        """
        query_embedding = cast(List[float], query.query_embedding)

        es_query = {}

        if query.filters is not None and len(query.filters.legacy_filters()) > 0:
            filter = [_to_elasticsearch_filter(query.filters)]
        else:
            filter = es_filter or []

        if query.mode in (
            VectorStoreQueryMode.DEFAULT,
            VectorStoreQueryMode.HYBRID,
        ):
            es_query["knn"] = {
                "filter": filter,
                "field": self.vector_field,
                "query_vector": query_embedding,
                "k": query.similarity_top_k,
                "num_candidates": query.similarity_top_k * 10,
            }

        if query.mode in (
            VectorStoreQueryMode.TEXT_SEARCH,
            VectorStoreQueryMode.HYBRID,
        ):
            es_query["query"] = {
                "bool": {
                    "must": {"match": {self.text_field: {"query": query.query_str}}},
                    "filter": filter,
                }
            }

        if query.mode == VectorStoreQueryMode.HYBRID:
            es_query["rank"] = {"rrf": {"window_size": query.similarity_top_k}}

        if custom_query is not None:
            es_query = custom_query(es_query, query)
            logger.debug(f"Calling custom_query, Query body now: {es_query}")

        async with self.client as client:
            response = await client.search(
                index=self.index_name,
                **es_query,
                size=query.similarity_top_k,
                _source={"excludes": [self.vector_field]},
            )

        top_k_nodes = []
        top_k_ids = []
        top_k_scores = []
        hits = response["hits"]["hits"]
        for hit in hits:
            source = hit["_source"]
            metadata = source.get("metadata", None)
            text = source.get(self.text_field, None)
            node_id = hit["_id"]

            try:
                node = metadata_dict_to_node(metadata)
                node.text = text
            except Exception:
                # Legacy support for old metadata format
                logger.warning(
                    f"Could not parse metadata from hit {hit['_source']['metadata']}"
                )
                node_info = source.get("node_info")
                relationships = source.get("relationships", {})
                start_char_idx = None
                end_char_idx = None
                if isinstance(node_info, dict):
                    start_char_idx = node_info.get("start", None)
                    end_char_idx = node_info.get("end", None)

                node = TextNode(
                    text=text,
                    metadata=metadata,
                    id_=node_id,
                    start_char_idx=start_char_idx,
                    end_char_idx=end_char_idx,
                    relationships=relationships,
                )
            top_k_nodes.append(node)
            top_k_ids.append(node_id)
            top_k_scores.append(hit.get("_rank", hit["_score"]))

        if query.mode == VectorStoreQueryMode.HYBRID:
            total_rank = sum(top_k_scores)
            top_k_scores = [total_rank - rank / total_rank for rank in top_k_scores]

        return VectorStoreQueryResult(
            nodes=top_k_nodes,
            ids=top_k_ids,
            similarities=_to_llama_similarities(top_k_scores),
        )

client property #

client: Any

Get async elasticsearch client.

get_user_agent staticmethod #

get_user_agent() -> str

Get user agent for elasticsearch client.

Source code in llama-index-integrations/vector_stores/llama-index-vector-stores-elasticsearch/llama_index/vector_stores/elasticsearch/base.py
264
265
266
267
268
269
270
@staticmethod
def get_user_agent() -> str:
    """Get user agent for elasticsearch client."""
    import llama_index.core

    version = getattr(llama_index.core, "__version__", "")
    return f"llama_index-py-vs/{version}"

add #

add(nodes: List[BaseNode], *, create_index_if_not_exists: bool = True, **add_kwargs: Any) -> List[str]

Add nodes to Elasticsearch index.

Parameters:

Name Type Description Default
nodes List[BaseNode]

List of nodes with embeddings.

required
create_index_if_not_exists bool

Optional. Whether to create the Elasticsearch index if it doesn't already exist. Defaults to True.

True

Returns:

Type Description
List[str]

List of node IDs that were added to the index.

Raises:

Type Description
ImportError

If elasticsearch['async'] python package is not installed.

BulkIndexError

If AsyncElasticsearch async_bulk indexing fails.

Source code in llama-index-integrations/vector_stores/llama-index-vector-stores-elasticsearch/llama_index/vector_stores/elasticsearch/base.py
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
def add(
    self,
    nodes: List[BaseNode],
    *,
    create_index_if_not_exists: bool = True,
    **add_kwargs: Any,
) -> List[str]:
    """Add nodes to Elasticsearch index.

    Args:
        nodes: List of nodes with embeddings.
        create_index_if_not_exists: Optional. Whether to create
                                    the Elasticsearch index if it
                                    doesn't already exist.
                                    Defaults to True.

    Returns:
        List of node IDs that were added to the index.

    Raises:
        ImportError: If elasticsearch['async'] python package is not installed.
        BulkIndexError: If AsyncElasticsearch async_bulk indexing fails.
    """
    return asyncio.get_event_loop().run_until_complete(
        self.async_add(nodes, create_index_if_not_exists=create_index_if_not_exists)
    )

async_add async #

async_add(nodes: List[BaseNode], *, create_index_if_not_exists: bool = True, **add_kwargs: Any) -> List[str]

Asynchronous method to add nodes to Elasticsearch index.

Parameters:

Name Type Description Default
nodes List[BaseNode]

List of nodes with embeddings.

required
create_index_if_not_exists bool

Optional. Whether to create the AsyncElasticsearch index if it doesn't already exist. Defaults to True.

True

Returns:

Type Description
List[str]

List of node IDs that were added to the index.

Raises:

Type Description
ImportError

If elasticsearch python package is not installed.

BulkIndexError

If AsyncElasticsearch async_bulk indexing fails.

Source code in llama-index-integrations/vector_stores/llama-index-vector-stores-elasticsearch/llama_index/vector_stores/elasticsearch/base.py
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
async def async_add(
    self,
    nodes: List[BaseNode],
    *,
    create_index_if_not_exists: bool = True,
    **add_kwargs: Any,
) -> List[str]:
    """Asynchronous method to add nodes to Elasticsearch index.

    Args:
        nodes: List of nodes with embeddings.
        create_index_if_not_exists: Optional. Whether to create
                                    the AsyncElasticsearch index if it
                                    doesn't already exist.
                                    Defaults to True.

    Returns:
        List of node IDs that were added to the index.

    Raises:
        ImportError: If elasticsearch python package is not installed.
        BulkIndexError: If AsyncElasticsearch async_bulk indexing fails.
    """
    if len(nodes) == 0:
        return []

    if create_index_if_not_exists:
        dims_length = len(nodes[0].get_embedding())
        await self._create_index_if_not_exists(
            index_name=self.index_name, dims_length=dims_length
        )

    embeddings: List[List[float]] = []
    texts: List[str] = []
    metadatas: List[dict] = []
    ids: List[str] = []
    for node in nodes:
        ids.append(node.node_id)
        embeddings.append(node.get_embedding())
        texts.append(node.get_content(metadata_mode=MetadataMode.NONE))
        metadatas.append(node_to_metadata_dict(node, remove_text=True))

    requests = []
    return_ids = []

    for i, text in enumerate(texts):
        metadata = metadatas[i] if metadatas else {}
        _id = ids[i] if ids else str(uuid.uuid4())
        request = {
            "_op_type": "index",
            "_index": self.index_name,
            self.vector_field: embeddings[i],
            self.text_field: text,
            "metadata": metadata,
            "_id": _id,
        }
        requests.append(request)
        return_ids.append(_id)

    async with self.client as client:
        await async_bulk(client, requests, chunk_size=self.batch_size, refresh=True)
        try:
            success, failed = await async_bulk(
                client, requests, stats_only=True, refresh=True
            )
            logger.debug(
                f"Added {success} and failed to add {failed} texts to index"
            )

            logger.debug(f"added texts {ids} to index")
            return return_ids
        except BulkIndexError as e:
            logger.error(f"Error adding texts: {e}")
            firstError = e.errors[0].get("index", {}).get("error", {})
            logger.error(f"First error reason: {firstError.get('reason')}")
            raise

delete #

delete(ref_doc_id: str, **delete_kwargs: Any) -> None

Delete node from Elasticsearch index.

Parameters:

Name Type Description Default
ref_doc_id str

ID of the node to delete.

required
delete_kwargs Any

Optional. Additional arguments to pass to Elasticsearch delete_by_query.

{}

Raises:

Type Description
Exception

If Elasticsearch delete_by_query fails.

Source code in llama-index-integrations/vector_stores/llama-index-vector-stores-elasticsearch/llama_index/vector_stores/elasticsearch/base.py
432
433
434
435
436
437
438
439
440
441
442
443
444
445
def delete(self, ref_doc_id: str, **delete_kwargs: Any) -> None:
    """Delete node from Elasticsearch index.

    Args:
        ref_doc_id: ID of the node to delete.
        delete_kwargs: Optional. Additional arguments to
                    pass to Elasticsearch delete_by_query.

    Raises:
        Exception: If Elasticsearch delete_by_query fails.
    """
    return asyncio.get_event_loop().run_until_complete(
        self.adelete(ref_doc_id, **delete_kwargs)
    )

adelete async #

adelete(ref_doc_id: str, **delete_kwargs: Any) -> None

Async delete node from Elasticsearch index.

Parameters:

Name Type Description Default
ref_doc_id str

ID of the node to delete.

required
delete_kwargs Any

Optional. Additional arguments to pass to AsyncElasticsearch delete_by_query.

{}

Raises:

Type Description
Exception

If AsyncElasticsearch delete_by_query fails.

Source code in llama-index-integrations/vector_stores/llama-index-vector-stores-elasticsearch/llama_index/vector_stores/elasticsearch/base.py
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
async def adelete(self, ref_doc_id: str, **delete_kwargs: Any) -> None:
    """Async delete node from Elasticsearch index.

    Args:
        ref_doc_id: ID of the node to delete.
        delete_kwargs: Optional. Additional arguments to
                    pass to AsyncElasticsearch delete_by_query.

    Raises:
        Exception: If AsyncElasticsearch delete_by_query fails.
    """
    try:
        async with self.client as client:
            res = await client.delete_by_query(
                index=self.index_name,
                query={"term": {"metadata.ref_doc_id": ref_doc_id}},
                refresh=True,
                **delete_kwargs,
            )
        if res["deleted"] == 0:
            logger.warning(f"Could not find text {ref_doc_id} to delete")
        else:
            logger.debug(f"Deleted text {ref_doc_id} from index")
    except Exception:
        logger.error(f"Error deleting text: {ref_doc_id}")
        raise

query #

query(query: VectorStoreQuery, custom_query: Optional[Callable[[Dict, Union[VectorStoreQuery, None]], Dict]] = None, es_filter: Optional[List[Dict]] = None, **kwargs: Any) -> VectorStoreQueryResult

Query index for top k most similar nodes.

Parameters:

Name Type Description Default
query_embedding List[float]

query embedding

required
custom_query Optional[Callable[[Dict, Union[VectorStoreQuery, None]], Dict]]

Optional. custom query function that takes in the es query body and returns a modified query body. This can be used to add additional query parameters to the Elasticsearch query.

None
es_filter Optional[List[Dict]]

Optional. Elasticsearch filter to apply to the query. If filter is provided in the query, this filter will be ignored.

None

Returns:

Name Type Description
VectorStoreQueryResult VectorStoreQueryResult

Result of the query.

Raises:

Type Description
Exception

If Elasticsearch query fails.

Source code in llama-index-integrations/vector_stores/llama-index-vector-stores-elasticsearch/llama_index/vector_stores/elasticsearch/base.py
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
def query(
    self,
    query: VectorStoreQuery,
    custom_query: Optional[
        Callable[[Dict, Union[VectorStoreQuery, None]], Dict]
    ] = None,
    es_filter: Optional[List[Dict]] = None,
    **kwargs: Any,
) -> VectorStoreQueryResult:
    """Query index for top k most similar nodes.

    Args:
        query_embedding (List[float]): query embedding
        custom_query: Optional. custom query function that takes in the es query
                    body and returns a modified query body.
                    This can be used to add additional query
                    parameters to the Elasticsearch query.
        es_filter: Optional. Elasticsearch filter to apply to the
                    query. If filter is provided in the query,
                    this filter will be ignored.

    Returns:
        VectorStoreQueryResult: Result of the query.

    Raises:
        Exception: If Elasticsearch query fails.

    """
    return asyncio.get_event_loop().run_until_complete(
        self.aquery(query, custom_query, es_filter, **kwargs)
    )

aquery async #

aquery(query: VectorStoreQuery, custom_query: Optional[Callable[[Dict, Union[VectorStoreQuery, None]], Dict]] = None, es_filter: Optional[List[Dict]] = None, **kwargs: Any) -> VectorStoreQueryResult

Asynchronous query index for top k most similar nodes.

Parameters:

Name Type Description Default
query_embedding VectorStoreQuery

query embedding

required
custom_query Optional[Callable[[Dict, Union[VectorStoreQuery, None]], Dict]]

Optional. custom query function that takes in the es query body and returns a modified query body. This can be used to add additional query parameters to the AsyncElasticsearch query.

None
es_filter Optional[List[Dict]]

Optional. AsyncElasticsearch filter to apply to the query. If filter is provided in the query, this filter will be ignored.

None

Returns:

Name Type Description
VectorStoreQueryResult VectorStoreQueryResult

Result of the query.

Raises:

Type Description
Exception

If AsyncElasticsearch query fails.

Source code in llama-index-integrations/vector_stores/llama-index-vector-stores-elasticsearch/llama_index/vector_stores/elasticsearch/base.py
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
async def aquery(
    self,
    query: VectorStoreQuery,
    custom_query: Optional[
        Callable[[Dict, Union[VectorStoreQuery, None]], Dict]
    ] = None,
    es_filter: Optional[List[Dict]] = None,
    **kwargs: Any,
) -> VectorStoreQueryResult:
    """Asynchronous query index for top k most similar nodes.

    Args:
        query_embedding (VectorStoreQuery): query embedding
        custom_query: Optional. custom query function that takes in the es query
                    body and returns a modified query body.
                    This can be used to add additional query
                    parameters to the AsyncElasticsearch query.
        es_filter: Optional. AsyncElasticsearch filter to apply to the
                    query. If filter is provided in the query,
                    this filter will be ignored.

    Returns:
        VectorStoreQueryResult: Result of the query.

    Raises:
        Exception: If AsyncElasticsearch query fails.

    """
    query_embedding = cast(List[float], query.query_embedding)

    es_query = {}

    if query.filters is not None and len(query.filters.legacy_filters()) > 0:
        filter = [_to_elasticsearch_filter(query.filters)]
    else:
        filter = es_filter or []

    if query.mode in (
        VectorStoreQueryMode.DEFAULT,
        VectorStoreQueryMode.HYBRID,
    ):
        es_query["knn"] = {
            "filter": filter,
            "field": self.vector_field,
            "query_vector": query_embedding,
            "k": query.similarity_top_k,
            "num_candidates": query.similarity_top_k * 10,
        }

    if query.mode in (
        VectorStoreQueryMode.TEXT_SEARCH,
        VectorStoreQueryMode.HYBRID,
    ):
        es_query["query"] = {
            "bool": {
                "must": {"match": {self.text_field: {"query": query.query_str}}},
                "filter": filter,
            }
        }

    if query.mode == VectorStoreQueryMode.HYBRID:
        es_query["rank"] = {"rrf": {"window_size": query.similarity_top_k}}

    if custom_query is not None:
        es_query = custom_query(es_query, query)
        logger.debug(f"Calling custom_query, Query body now: {es_query}")

    async with self.client as client:
        response = await client.search(
            index=self.index_name,
            **es_query,
            size=query.similarity_top_k,
            _source={"excludes": [self.vector_field]},
        )

    top_k_nodes = []
    top_k_ids = []
    top_k_scores = []
    hits = response["hits"]["hits"]
    for hit in hits:
        source = hit["_source"]
        metadata = source.get("metadata", None)
        text = source.get(self.text_field, None)
        node_id = hit["_id"]

        try:
            node = metadata_dict_to_node(metadata)
            node.text = text
        except Exception:
            # Legacy support for old metadata format
            logger.warning(
                f"Could not parse metadata from hit {hit['_source']['metadata']}"
            )
            node_info = source.get("node_info")
            relationships = source.get("relationships", {})
            start_char_idx = None
            end_char_idx = None
            if isinstance(node_info, dict):
                start_char_idx = node_info.get("start", None)
                end_char_idx = node_info.get("end", None)

            node = TextNode(
                text=text,
                metadata=metadata,
                id_=node_id,
                start_char_idx=start_char_idx,
                end_char_idx=end_char_idx,
                relationships=relationships,
            )
        top_k_nodes.append(node)
        top_k_ids.append(node_id)
        top_k_scores.append(hit.get("_rank", hit["_score"]))

    if query.mode == VectorStoreQueryMode.HYBRID:
        total_rank = sum(top_k_scores)
        top_k_scores = [total_rank - rank / total_rank for rank in top_k_scores]

    return VectorStoreQueryResult(
        nodes=top_k_nodes,
        ids=top_k_ids,
        similarities=_to_llama_similarities(top_k_scores),
    )