Struct ANNIvfSubIndexExec 

pub struct ANNIvfSubIndexExec { /* private fields */ }

Datafusion ExecutionPlan to run search on vector index partitions.

A IVF-{PQ/SQ/HNSW} query plan is:

AnnSubIndexExec: k=10
  AnnPartitionExec: nprobes=20

The partition index returns one batch per delta with maximum_nprobes partitions in sorted order.

The sub-index then runs a KNN search on each partition in parallel.

First, the index will search minimum_probes partitions on each delta. If there are enough results at that point to satisfy K then the results will be sorted and returned.

If there are not enough results then the prefilter will be consulted to determine how many potential results exist. If the number is smaller than K then those additional results will be fetched directly and given maximum distance.

If the number of results is larger then additional partitions will be searched in batches of cpu_parallelism until min(K, num_filtered_results) are found or maximum_nprobes partitions have been searched.

TODO: In the future, if we can know that a filter will be highly selective (through cost estimation or user-provided hints) we wait for the prefilter results before we load any partitions. If there are less than K (or some threshold) results then we can return without search.

Implementations

impl ANNIvfSubIndexExec

pub fn try_new( input: Arc<dyn ExecutionPlan>, dataset: Arc<Dataset>, indices: Vec<IndexMetadata>, query: Query, prefilter_source: PreFilterSource, ) -> Result<Self>

Trait Implementations

impl Debug for ANNIvfSubIndexExec

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl DisplayAs for ANNIvfSubIndexExec

fn fmt_as(&self, t: DisplayFormatType, f: &mut Formatter<'_>) -> Result

Format according to DisplayFormatType, used when verbose representation looks different from the default one

impl ExecutionPlan for ANNIvfSubIndexExec

fn name(&self) -> &str

Short name for the ExecutionPlan, such as ‘DataSourceExec’.

fn as_any(&self) -> &dyn Any

Returns the execution plan as Any so that it can be downcast to a specific implementation.

fn schema(&self) -> SchemaRef

Get the schema for this execution plan

fn children(&self) -> Vec<&Arc<dyn ExecutionPlan>>

Get a list of children ExecutionPlans that act as inputs to this plan. The returned list will be empty for leaf nodes such as scans, will contain a single value for unary nodes, or two values for binary nodes (such as joins).

fn required_input_distribution(&self) -> Vec<Distribution>

Specifies the data distribution requirements for all the children for this ExecutionPlan, By default it’s [Distribution::UnspecifiedDistribution] for each child,

fn with_new_children( self: Arc<Self>, children: Vec<Arc<dyn ExecutionPlan>>, ) -> DataFusionResult<Arc<dyn ExecutionPlan>>

Returns a new ExecutionPlan where all existing children were replaced by the children, in order

fn execute( &self, partition: usize, context: Arc<TaskContext>, ) -> DataFusionResult<SendableRecordBatchStream>

Begin execution of partition, returning a Stream of RecordBatches.

fn partition_statistics( &self, partition: Option<usize>, ) -> DataFusionResult<Statistics>

Returns statistics for a specific partition of this ExecutionPlan node. If statistics are not available, should return Statistics::new_unknown (the default), not an error. If partition is None, it returns statistics for the entire plan.

fn metrics(&self) -> Option<MetricsSet>

Return a snapshot of the set of Metrics for this ExecutionPlan. If no Metrics are available, return None.

fn properties(&self) -> &PlanProperties

Return properties of the output of the ExecutionPlan, such as output ordering(s), partitioning information etc.

fn supports_limit_pushdown(&self) -> bool

Returns true if a limit can be safely pushed down through this ExecutionPlan node.

fn static_name() -> &'static str

where Self: Sized,

Short name for the ExecutionPlan, such as ‘DataSourceExec’. Like name but can be called without an instance.

fn check_invariants(&self, check: InvariantLevel) -> Result<(), DataFusionError>

Returns an error if this individual node does not conform to its invariants. These invariants are typically only checked in debug mode.

fn required_input_ordering(&self) -> Vec<Option<OrderingRequirements>>

Specifies the ordering required for all of the children of this ExecutionPlan.

fn maintains_input_order(&self) -> Vec<bool>

Returns false if this ExecutionPlan’s implementation may reorder rows within or between partitions.

fn benefits_from_input_partitioning(&self) -> Vec<bool>

Specifies whether the ExecutionPlan benefits from increased parallelization at its input for each child.

fn reset_state( self: Arc<Self>, ) -> Result<Arc<dyn ExecutionPlan>, DataFusionError>

Reset any internal state within this ExecutionPlan.

fn repartitioned( &self, _target_partitions: usize, _config: &ConfigOptions, ) -> Result<Option<Arc<dyn ExecutionPlan>>, DataFusionError>

If supported, attempt to increase the partitioning of this ExecutionPlan to produce target_partitions partitions.

fn statistics(&self) -> Result<Statistics, DataFusionError>

👎Deprecated since 48.0.0: Use partition_statistics method instead

Returns statistics for this ExecutionPlan node. If statistics are not available, should return Statistics::new_unknown (the default), not an error.

fn with_fetch(&self, _limit: Option<usize>) -> Option<Arc<dyn ExecutionPlan>>

Returns a fetching variant of this ExecutionPlan node, if it supports fetch limits. Returns None otherwise.

fn fetch(&self) -> Option<usize>

Gets the fetch count for the operator, None means there is no fetch.

fn cardinality_effect(&self) -> CardinalityEffect

Gets the effect on cardinality, if known

fn try_swapping_with_projection( &self, _projection: &ProjectionExec, ) -> Result<Option<Arc<dyn ExecutionPlan>>, DataFusionError>

Attempts to push down the given projection into the input of this ExecutionPlan.

fn gather_filters_for_pushdown( &self, _phase: FilterPushdownPhase, parent_filters: Vec<Arc<dyn PhysicalExpr>>, _config: &ConfigOptions, ) -> Result<FilterDescription, DataFusionError>

Collect filters that this node can push down to its children. Filters that are being pushed down from parents are passed in, and the node may generate additional filters to push down. For example, given the plan FilterExec -> HashJoinExec -> DataSourceExec, what will happen is that we recurse down the plan calling ExecutionPlan::gather_filters_for_pushdown:

fn handle_child_pushdown_result( &self, _phase: FilterPushdownPhase, child_pushdown_result: ChildPushdownResult, _config: &ConfigOptions, ) -> Result<FilterPushdownPropagation<Arc<dyn ExecutionPlan>>, DataFusionError>

Handle the result of a child pushdown. This method is called as we recurse back up the plan tree after pushing filters down to child nodes via ExecutionPlan::gather_filters_for_pushdown. It allows the current node to process the results of filter pushdown from its children, deciding whether to absorb filters, modify the plan, or pass filters back up to its parent.

fn with_new_state( &self, _state: Arc<dyn Any + Send + Sync>, ) -> Option<Arc<dyn ExecutionPlan>>

Injects arbitrary run-time state into this execution plan, returning a new plan instance that incorporates that state if it is relevant to the concrete node implementation.