Index

Vector store index types.

VectorStoreQueryResult dataclass

Vector store query result.

Source code in llama-index-core/llama_index/core/vector_stores/types.py

@dataclass
class VectorStoreQueryResult:
    """Vector store query result."""

    nodes: Optional[Sequence[BaseNode]] = None
    similarities: Optional[List[float]] = None
    ids: Optional[List[str]] = None

VectorStoreQueryMode

Bases: str, Enum

Vector store query mode.

Source code in llama-index-core/llama_index/core/vector_stores/types.py

class VectorStoreQueryMode(str, Enum):
    """Vector store query mode."""

    DEFAULT = "default"
    SPARSE = "sparse"
    HYBRID = "hybrid"
    TEXT_SEARCH = "text_search"
    SEMANTIC_HYBRID = "semantic_hybrid"

    # fit learners
    SVM = "svm"
    LOGISTIC_REGRESSION = "logistic_regression"
    LINEAR_REGRESSION = "linear_regression"

    # maximum marginal relevance
    MMR = "mmr"

FilterOperator

Bases: str, Enum

Vector store filter operator.

Source code in llama-index-core/llama_index/core/vector_stores/types.py

class FilterOperator(str, Enum):
    """Vector store filter operator."""

    # TODO add more operators
    EQ = "=="  # default operator (string, int, float)
    GT = ">"  # greater than (int, float)
    LT = "<"  # less than (int, float)
    NE = "!="  # not equal to (string, int, float)
    GTE = ">="  # greater than or equal to (int, float)
    LTE = "<="  # less than or equal to (int, float)
    IN = "in"  # metadata in value array (string or number)
    NIN = "nin"  # metadata not in value array (string or number)
    TEXT_MATCH = "text_match"  # full text match (allows you to search for a specific substring, token or phrase within the text field)
    CONTAINS = "contains"  # metadata array contains value (string or number)

FilterCondition

Bases: str, Enum

Vector store filter conditions to combine different filters.

Source code in llama-index-core/llama_index/core/vector_stores/types.py

class FilterCondition(str, Enum):
    """Vector store filter conditions to combine different filters."""

    # TODO add more conditions
    AND = "and"
    OR = "or"

MetadataFilter

Bases: BaseModel

Comprehensive metadata filter for vector stores to support more operators.

Value uses Strict* types, as int, float and str are compatible types and were all converted to string before.

See: https://docs.pydantic.dev/latest/usage/types/#strict-types

Source code in llama-index-core/llama_index/core/vector_stores/types.py

class MetadataFilter(BaseModel):
    """Comprehensive metadata filter for vector stores to support more operators.

    Value uses Strict* types, as int, float and str are compatible types and were all
    converted to string before.

    See: https://docs.pydantic.dev/latest/usage/types/#strict-types
    """

    key: str
    value: Union[
        StrictInt,
        StrictFloat,
        StrictStr,
        List[Union[StrictInt, StrictFloat, StrictStr]],
    ]
    operator: FilterOperator = FilterOperator.EQ

    @classmethod
    def from_dict(
        cls,
        filter_dict: Dict,
    ) -> "MetadataFilter":
        """Create MetadataFilter from dictionary.

        Args:
            filter_dict: Dict with key, value and operator.

        """
        return MetadataFilter.parse_obj(filter_dict)

from_dict classmethod

from_dict(filter_dict: Dict) -> MetadataFilter

Create MetadataFilter from dictionary.

Parameters:

Name	Type	Description	Default
filter_dict	Dict	Dict with key, value and operator.	required

Source code in llama-index-core/llama_index/core/vector_stores/types.py

@classmethod
def from_dict(
    cls,
    filter_dict: Dict,
) -> "MetadataFilter":
    """Create MetadataFilter from dictionary.

    Args:
        filter_dict: Dict with key, value and operator.

    """
    return MetadataFilter.parse_obj(filter_dict)

MetadataFilters

Bases: BaseModel

Metadata filters for vector stores.

Source code in llama-index-core/llama_index/core/vector_stores/types.py

class MetadataFilters(BaseModel):
    """Metadata filters for vector stores."""

    # Exact match filters and Advanced filters with operators like >, <, >=, <=, !=, etc.
    filters: List[Union[MetadataFilter, ExactMatchFilter, "MetadataFilters"]]
    # and/or such conditions for combining different filters
    condition: Optional[FilterCondition] = FilterCondition.AND

    @classmethod
    @deprecated(
        "`from_dict()` is deprecated. "
        "Please use `MetadataFilters(filters=.., condition='and')` directly instead."
    )
    def from_dict(cls, filter_dict: Dict) -> "MetadataFilters":
        """Create MetadataFilters from json."""
        filters = []
        for k, v in filter_dict.items():
            filter = MetadataFilter(key=k, value=v, operator=FilterOperator.EQ)
            filters.append(filter)
        return cls(filters=filters)

    @classmethod
    def from_dicts(
        cls,
        filter_dicts: List[Dict],
        condition: Optional[FilterCondition] = FilterCondition.AND,
    ) -> "MetadataFilters":
        """Create MetadataFilters from dicts.

        This takes in a list of individual MetadataFilter objects, along
        with the condition.

        Args:
            filter_dicts: List of dicts, each dict is a MetadataFilter.
            condition: FilterCondition to combine different filters.

        """
        return cls(
            filters=[
                MetadataFilter.from_dict(filter_dict) for filter_dict in filter_dicts
            ],
            condition=condition,
        )

    def legacy_filters(self) -> List[ExactMatchFilter]:
        """Convert MetadataFilters to legacy ExactMatchFilters."""
        filters = []
        for filter in self.filters:
            if filter.operator != FilterOperator.EQ:
                raise ValueError(
                    "Vector Store only supports exact match filters. "
                    "Please use ExactMatchFilter or FilterOperator.EQ instead."
                )
            filters.append(ExactMatchFilter(key=filter.key, value=filter.value))
        return filters

from_dict classmethod

from_dict(filter_dict: Dict) -> MetadataFilters

Create MetadataFilters from json.

Source code in llama-index-core/llama_index/core/vector_stores/types.py

@classmethod
@deprecated(
    "`from_dict()` is deprecated. "
    "Please use `MetadataFilters(filters=.., condition='and')` directly instead."
)
def from_dict(cls, filter_dict: Dict) -> "MetadataFilters":
    """Create MetadataFilters from json."""
    filters = []
    for k, v in filter_dict.items():
        filter = MetadataFilter(key=k, value=v, operator=FilterOperator.EQ)
        filters.append(filter)
    return cls(filters=filters)

from_dicts classmethod

from_dicts(filter_dicts: List[Dict], condition: Optional[FilterCondition] = FilterCondition.AND) -> MetadataFilters

Create MetadataFilters from dicts.

This takes in a list of individual MetadataFilter objects, along with the condition.

Parameters:

Name	Type	Description	Default
filter_dicts	List[Dict]	List of dicts, each dict is a MetadataFilter.	required
condition	Optional[FilterCondition]	FilterCondition to combine different filters.	AND

Source code in llama-index-core/llama_index/core/vector_stores/types.py

@classmethod
def from_dicts(
    cls,
    filter_dicts: List[Dict],
    condition: Optional[FilterCondition] = FilterCondition.AND,
) -> "MetadataFilters":
    """Create MetadataFilters from dicts.

    This takes in a list of individual MetadataFilter objects, along
    with the condition.

    Args:
        filter_dicts: List of dicts, each dict is a MetadataFilter.
        condition: FilterCondition to combine different filters.

    """
    return cls(
        filters=[
            MetadataFilter.from_dict(filter_dict) for filter_dict in filter_dicts
        ],
        condition=condition,
    )

legacy_filters

legacy_filters() -> List[ExactMatchFilter]

Convert MetadataFilters to legacy ExactMatchFilters.

Source code in llama-index-core/llama_index/core/vector_stores/types.py

def legacy_filters(self) -> List[ExactMatchFilter]:
    """Convert MetadataFilters to legacy ExactMatchFilters."""
    filters = []
    for filter in self.filters:
        if filter.operator != FilterOperator.EQ:
            raise ValueError(
                "Vector Store only supports exact match filters. "
                "Please use ExactMatchFilter or FilterOperator.EQ instead."
            )
        filters.append(ExactMatchFilter(key=filter.key, value=filter.value))
    return filters

VectorStoreQuerySpec

Bases: BaseModel

Schema for a structured request for vector store (i.e. to be converted to a VectorStoreQuery).

Currently only used by VectorIndexAutoRetriever.

Source code in llama-index-core/llama_index/core/vector_stores/types.py

class VectorStoreQuerySpec(BaseModel):
    """Schema for a structured request for vector store
    (i.e. to be converted to a VectorStoreQuery).

    Currently only used by VectorIndexAutoRetriever.
    """

    query: str
    filters: List[MetadataFilter]
    top_k: Optional[int] = None

MetadataInfo

Bases: BaseModel

Information about a metadata filter supported by a vector store.

Currently only used by VectorIndexAutoRetriever.

Source code in llama-index-core/llama_index/core/vector_stores/types.py

class MetadataInfo(BaseModel):
    """Information about a metadata filter supported by a vector store.

    Currently only used by VectorIndexAutoRetriever.
    """

    name: str
    type: str
    description: str

VectorStoreInfo

Bases: BaseModel

Information about a vector store (content and supported metadata filters).

Currently only used by VectorIndexAutoRetriever.

Source code in llama-index-core/llama_index/core/vector_stores/types.py

class VectorStoreInfo(BaseModel):
    """Information about a vector store (content and supported metadata filters).

    Currently only used by VectorIndexAutoRetriever.
    """

    metadata_info: List[MetadataInfo]
    content_info: str

VectorStoreQuery dataclass

Vector store query.

Source code in llama-index-core/llama_index/core/vector_stores/types.py

@dataclass
class VectorStoreQuery:
    """Vector store query."""

    query_embedding: Optional[List[float]] = None
    similarity_top_k: int = 1
    doc_ids: Optional[List[str]] = None
    node_ids: Optional[List[str]] = None
    query_str: Optional[str] = None
    output_fields: Optional[List[str]] = None
    embedding_field: Optional[str] = None

    mode: VectorStoreQueryMode = VectorStoreQueryMode.DEFAULT

    # NOTE: only for hybrid search (0 for bm25, 1 for vector search)
    alpha: Optional[float] = None

    # metadata filters
    filters: Optional[MetadataFilters] = None

    # only for mmr
    mmr_threshold: Optional[float] = None

    # NOTE: currently only used by postgres hybrid search
    sparse_top_k: Optional[int] = None
    # NOTE: return top k results from hybrid search. similarity_top_k is used for dense search top k
    hybrid_top_k: Optional[int] = None

VectorStore

Bases: Protocol

Abstract vector store protocol.

Source code in llama-index-core/llama_index/core/vector_stores/types.py

@runtime_checkable
class VectorStore(Protocol):
    """Abstract vector store protocol."""

    stores_text: bool
    is_embedding_query: bool = True

    @property
    def client(self) -> Any:
        """Get client."""
        ...

    def add(
        self,
        nodes: List[BaseNode],
        **add_kwargs: Any,
    ) -> List[str]:
        """Add nodes with embedding to vector store."""
        ...

    async def async_add(
        self,
        nodes: List[BaseNode],
        **kwargs: Any,
    ) -> List[str]:
        """
        Asynchronously add nodes with embedding to vector store.
        NOTE: this is not implemented for all vector stores. If not implemented,
        it will just call add synchronously.
        """
        return self.add(nodes)

    def delete(self, ref_doc_id: str, **delete_kwargs: Any) -> None:
        """
        Delete nodes using with ref_doc_id."""
        ...

    async def adelete(self, ref_doc_id: str, **delete_kwargs: Any) -> None:
        """
        Delete nodes using with ref_doc_id.
        NOTE: this is not implemented for all vector stores. If not implemented,
        it will just call delete synchronously.
        """
        self.delete(ref_doc_id, **delete_kwargs)

    def query(self, query: VectorStoreQuery, **kwargs: Any) -> VectorStoreQueryResult:
        """Query vector store."""
        ...

    async def aquery(
        self, query: VectorStoreQuery, **kwargs: Any
    ) -> VectorStoreQueryResult:
        """
        Asynchronously query vector store.
        NOTE: this is not implemented for all vector stores. If not implemented,
        it will just call query synchronously.
        """
        return self.query(query, **kwargs)

    def persist(
        self, persist_path: str, fs: Optional[fsspec.AbstractFileSystem] = None
    ) -> None:
        return None

client property

client: Any

Get client.

add

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

Add nodes with embedding to vector store.

Source code in llama-index-core/llama_index/core/vector_stores/types.py

def add(
    self,
    nodes: List[BaseNode],
    **add_kwargs: Any,
) -> List[str]:
    """Add nodes with embedding to vector store."""
    ...

async_add async

async_add(nodes: List[BaseNode], **kwargs: Any) -> List[str]

Asynchronously add nodes with embedding to vector store. NOTE: this is not implemented for all vector stores. If not implemented, it will just call add synchronously.

Source code in llama-index-core/llama_index/core/vector_stores/types.py

async def async_add(
    self,
    nodes: List[BaseNode],
    **kwargs: Any,
) -> List[str]:
    """
    Asynchronously add nodes with embedding to vector store.
    NOTE: this is not implemented for all vector stores. If not implemented,
    it will just call add synchronously.
    """
    return self.add(nodes)

delete

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

Delete nodes using with ref_doc_id.

Source code in llama-index-core/llama_index/core/vector_stores/types.py

def delete(self, ref_doc_id: str, **delete_kwargs: Any) -> None:
    """
    Delete nodes using with ref_doc_id."""
    ...

adelete async

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

Delete nodes using with ref_doc_id. NOTE: this is not implemented for all vector stores. If not implemented, it will just call delete synchronously.

Source code in llama-index-core/llama_index/core/vector_stores/types.py

async def adelete(self, ref_doc_id: str, **delete_kwargs: Any) -> None:
    """
    Delete nodes using with ref_doc_id.
    NOTE: this is not implemented for all vector stores. If not implemented,
    it will just call delete synchronously.
    """
    self.delete(ref_doc_id, **delete_kwargs)

query

query(query: VectorStoreQuery, **kwargs: Any) -> VectorStoreQueryResult

Query vector store.

Source code in llama-index-core/llama_index/core/vector_stores/types.py

def query(self, query: VectorStoreQuery, **kwargs: Any) -> VectorStoreQueryResult:
    """Query vector store."""
    ...

aquery async

aquery(query: VectorStoreQuery, **kwargs: Any) -> VectorStoreQueryResult

Asynchronously query vector store. NOTE: this is not implemented for all vector stores. If not implemented, it will just call query synchronously.

Source code in llama-index-core/llama_index/core/vector_stores/types.py

async def aquery(
    self, query: VectorStoreQuery, **kwargs: Any
) -> VectorStoreQueryResult:
    """
    Asynchronously query vector store.
    NOTE: this is not implemented for all vector stores. If not implemented,
    it will just call query synchronously.
    """
    return self.query(query, **kwargs)

BasePydanticVectorStore

Bases: BaseComponent, ABC

Abstract vector store protocol.

Source code in llama-index-core/llama_index/core/vector_stores/types.py

class BasePydanticVectorStore(BaseComponent, ABC):
    """Abstract vector store protocol."""

    stores_text: bool
    is_embedding_query: bool = True

    @property
    @abstractmethod
    def client(self) -> Any:
        """Get client."""

    @abstractmethod
    def add(
        self,
        nodes: List[BaseNode],
    ) -> List[str]:
        """Add nodes to vector store."""

    async def async_add(
        self,
        nodes: List[BaseNode],
        **kwargs: Any,
    ) -> List[str]:
        """
        Asynchronously add nodes to vector store.
        NOTE: this is not implemented for all vector stores. If not implemented,
        it will just call add synchronously.
        """
        return self.add(nodes)

    @abstractmethod
    def delete(self, ref_doc_id: str, **delete_kwargs: Any) -> None:
        """
        Delete nodes using with ref_doc_id."""

    async def adelete(self, ref_doc_id: str, **delete_kwargs: Any) -> None:
        """
        Delete nodes using with ref_doc_id.
        NOTE: this is not implemented for all vector stores. If not implemented,
        it will just call delete synchronously.
        """
        self.delete(ref_doc_id, **delete_kwargs)

    @abstractmethod
    def query(self, query: VectorStoreQuery, **kwargs: Any) -> VectorStoreQueryResult:
        """Query vector store."""

    async def aquery(
        self, query: VectorStoreQuery, **kwargs: Any
    ) -> VectorStoreQueryResult:
        """
        Asynchronously query vector store.
        NOTE: this is not implemented for all vector stores. If not implemented,
        it will just call query synchronously.
        """
        return self.query(query, **kwargs)

    def persist(
        self, persist_path: str, fs: Optional[fsspec.AbstractFileSystem] = None
    ) -> None:
        return None

client abstractmethod property

client: Any

Get client.

add abstractmethod

add(nodes: List[BaseNode]) -> List[str]

Add nodes to vector store.

Source code in llama-index-core/llama_index/core/vector_stores/types.py

@abstractmethod
def add(
    self,
    nodes: List[BaseNode],
) -> List[str]:
    """Add nodes to vector store."""

async_add async

async_add(nodes: List[BaseNode], **kwargs: Any) -> List[str]

Asynchronously add nodes to vector store. NOTE: this is not implemented for all vector stores. If not implemented, it will just call add synchronously.

Source code in llama-index-core/llama_index/core/vector_stores/types.py

async def async_add(
    self,
    nodes: List[BaseNode],
    **kwargs: Any,
) -> List[str]:
    """
    Asynchronously add nodes to vector store.
    NOTE: this is not implemented for all vector stores. If not implemented,
    it will just call add synchronously.
    """
    return self.add(nodes)

delete abstractmethod

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

Delete nodes using with ref_doc_id.

Source code in llama-index-core/llama_index/core/vector_stores/types.py

@abstractmethod
def delete(self, ref_doc_id: str, **delete_kwargs: Any) -> None:
    """
    Delete nodes using with ref_doc_id."""

adelete async

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

Delete nodes using with ref_doc_id. NOTE: this is not implemented for all vector stores. If not implemented, it will just call delete synchronously.

Source code in llama-index-core/llama_index/core/vector_stores/types.py

async def adelete(self, ref_doc_id: str, **delete_kwargs: Any) -> None:
    """
    Delete nodes using with ref_doc_id.
    NOTE: this is not implemented for all vector stores. If not implemented,
    it will just call delete synchronously.
    """
    self.delete(ref_doc_id, **delete_kwargs)

query abstractmethod

query(query: VectorStoreQuery, **kwargs: Any) -> VectorStoreQueryResult

Query vector store.

Source code in llama-index-core/llama_index/core/vector_stores/types.py

@abstractmethod
def query(self, query: VectorStoreQuery, **kwargs: Any) -> VectorStoreQueryResult:
    """Query vector store."""

aquery async

aquery(query: VectorStoreQuery, **kwargs: Any) -> VectorStoreQueryResult

Asynchronously query vector store. NOTE: this is not implemented for all vector stores. If not implemented, it will just call query synchronously.

Source code in llama-index-core/llama_index/core/vector_stores/types.py

async def aquery(
    self, query: VectorStoreQuery, **kwargs: Any
) -> VectorStoreQueryResult:
    """
    Asynchronously query vector store.
    NOTE: this is not implemented for all vector stores. If not implemented,
    it will just call query synchronously.
    """
    return self.query(query, **kwargs)