org.apache.hadoop.hive.ql.exec.vector.expressions.aggregates

Class VectorAggregateExpression

java.lang.Object

org.apache.hadoop.hive.ql.exec.vector.expressions.aggregates.VectorAggregateExpression

All Implemented Interfaces:

Serializable

Direct Known Subclasses:

VectorUDAFAvgDecimal, VectorUDAFAvgDecimal64ToDecimal, VectorUDAFAvgDecimal64ToDecimalComplete, VectorUDAFAvgDecimalComplete, VectorUDAFAvgDecimalFinal, VectorUDAFAvgDecimalPartial2, VectorUDAFAvgDouble, VectorUDAFAvgDoubleComplete, VectorUDAFAvgFinal, VectorUDAFAvgLong, VectorUDAFAvgLongComplete, VectorUDAFAvgPartial2, VectorUDAFAvgTimestamp, VectorUDAFAvgTimestampComplete, VectorUDAFBloomFilter, VectorUDAFBloomFilterMerge, VectorUDAFComputeBitVectorDecimal, VectorUDAFComputeBitVectorDouble, VectorUDAFComputeBitVectorFinal, VectorUDAFComputeBitVectorLong, VectorUDAFComputeBitVectorString, VectorUDAFComputeBitVectorTimestamp, VectorUDAFComputeDsKllSketchDouble, VectorUDAFComputeDsKllSketchFinal, VectorUDAFCount, VectorUDAFCountMerge, VectorUDAFCountStar, VectorUDAFMaxDecimal, VectorUDAFMaxDouble, VectorUDAFMaxIntervalDayTime, VectorUDAFMaxLong, VectorUDAFMaxString, VectorUDAFMaxTimestamp, VectorUDAFMinDecimal, VectorUDAFMinDouble, VectorUDAFMinIntervalDayTime, VectorUDAFMinLong, VectorUDAFMinString, VectorUDAFMinTimestamp, VectorUDAFSumDecimal, VectorUDAFSumDecimal64, VectorUDAFSumDecimal64ToDecimal, VectorUDAFSumDouble, VectorUDAFSumLong, VectorUDAFSumTimestamp, VectorUDAFVarDecimal, VectorUDAFVarDecimalComplete, VectorUDAFVarDouble, VectorUDAFVarDoubleComplete, VectorUDAFVarFinal, VectorUDAFVarLong, VectorUDAFVarLongComplete, VectorUDAFVarPartial2, VectorUDAFVarTimestamp, VectorUDAFVarTimestampComplete

public abstract class VectorAggregateExpression
extends Object
implements Serializable

Base class for aggregation expressions.

See Also:

Serialized Form

Nested Class Summary

Nested Classes

Modifier and Type	Class and Description
static interface	VectorAggregateExpression.AggregationBuffer Buffer interface to store aggregates.

Field Summary

Fields

Modifier and Type	Field and Description
static int	AVERAGE_COUNT_FIELD_INDEX
static int	AVERAGE_SOURCE_FIELD_INDEX
static int	AVERAGE_SUM_FIELD_INDEX
protected VectorExpression	inputExpression
protected TypeInfo	inputTypeInfo
protected GenericUDAFEvaluator.Mode	mode
protected DataTypePhysicalVariation	outputDataTypePhysicalVariation
protected TypeInfo	outputTypeInfo
static int	VARIANCE_COUNT_FIELD_INDEX
static int	VARIANCE_SUM_FIELD_INDEX
static int	VARIANCE_VARIANCE_FIELD_INDEX
protected VectorAggregationDesc	vecAggrDesc

Constructor Summary

Constructors

Constructor and Description
VectorAggregateExpression()
VectorAggregateExpression(VectorAggregationDesc vecAggrDesc)

Method Summary

Modifier and Type	Method and Description
abstract void	aggregateInput(VectorAggregateExpression.AggregationBuffer agg, VectorizedRowBatch unit)
abstract void	aggregateInputSelection(VectorAggregationBufferRow[] aggregationBufferSets, int aggregateIndex, VectorizedRowBatch vrg)
abstract void	assignRowColumn(VectorizedRowBatch batch, int batchIndex, int columnNum, VectorAggregateExpression.AggregationBuffer agg)
boolean	batchNeedsClone() This method should return true, when the particular VectorAggregateExpression must use a cloned batch while processing one, see Operator.batchNeedsClone() for further details.
void	finish(VectorAggregateExpression.AggregationBuffer aggregationBuffer, boolean aborted) Optional method to implement in VectorAggregateExpression instances.
abstract long	getAggregationBufferFixedSize()
VectorExpression	getInputExpression()
abstract VectorAggregateExpression.AggregationBuffer	getNewAggregationBuffer()
DataTypePhysicalVariation	getOutputDataTypePhysicalVariation()
TypeInfo	getOutputTypeInfo()
boolean	hasVariableSize()
abstract boolean	matches(String name, ColumnVector.Type inputColVectorType, ColumnVector.Type outputColVectorType, GenericUDAFEvaluator.Mode mode)
abstract void	reset(VectorAggregateExpression.AggregationBuffer agg)
String	toString()

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, wait, wait, wait

Field Detail

vecAggrDesc

protected final VectorAggregationDesc vecAggrDesc

inputExpression

protected final VectorExpression inputExpression

inputTypeInfo

protected final TypeInfo inputTypeInfo

outputTypeInfo

protected final TypeInfo outputTypeInfo

outputDataTypePhysicalVariation

protected final DataTypePhysicalVariation outputDataTypePhysicalVariation

mode

protected final GenericUDAFEvaluator.Mode mode

AVERAGE_COUNT_FIELD_INDEX

public static final int AVERAGE_COUNT_FIELD_INDEX

See Also:

Constant Field Values

AVERAGE_SUM_FIELD_INDEX

public static final int AVERAGE_SUM_FIELD_INDEX

See Also:

Constant Field Values

AVERAGE_SOURCE_FIELD_INDEX

public static final int AVERAGE_SOURCE_FIELD_INDEX

See Also:

Constant Field Values

VARIANCE_COUNT_FIELD_INDEX

public static final int VARIANCE_COUNT_FIELD_INDEX

See Also:

Constant Field Values

VARIANCE_SUM_FIELD_INDEX

public static final int VARIANCE_SUM_FIELD_INDEX

See Also:

Constant Field Values

VARIANCE_VARIANCE_FIELD_INDEX

public static final int VARIANCE_VARIANCE_FIELD_INDEX

See Also:

Constant Field Values

Constructor Detail

VectorAggregateExpression

public VectorAggregateExpression()

VectorAggregateExpression

public VectorAggregateExpression(VectorAggregationDesc vecAggrDesc)

Method Detail

getInputExpression

public VectorExpression getInputExpression()

getOutputTypeInfo

public TypeInfo getOutputTypeInfo()

getOutputDataTypePhysicalVariation

public DataTypePhysicalVariation getOutputDataTypePhysicalVariation()

getNewAggregationBuffer

public abstract VectorAggregateExpression.AggregationBuffer getNewAggregationBuffer()
                                                                             throws HiveException

Throws:

HiveException

aggregateInput

public abstract void aggregateInput(VectorAggregateExpression.AggregationBuffer agg,
                                    VectorizedRowBatch unit)
                             throws HiveException

Throws:

HiveException

aggregateInputSelection

public abstract void aggregateInputSelection(VectorAggregationBufferRow[] aggregationBufferSets,
                                             int aggregateIndex,
                                             VectorizedRowBatch vrg)
                                      throws HiveException

Throws:

HiveException

reset

public abstract void reset(VectorAggregateExpression.AggregationBuffer agg)
                    throws HiveException

Throws:

HiveException

getAggregationBufferFixedSize

public abstract long getAggregationBufferFixedSize()

hasVariableSize

public boolean hasVariableSize()

matches

public abstract boolean matches(String name,
                                ColumnVector.Type inputColVectorType,
                                ColumnVector.Type outputColVectorType,
                                GenericUDAFEvaluator.Mode mode)

assignRowColumn

public abstract void assignRowColumn(VectorizedRowBatch batch,
                                     int batchIndex,
                                     int columnNum,
                                     VectorAggregateExpression.AggregationBuffer agg)
                              throws HiveException

Throws:

HiveException

toString

public String toString()

Overrides:

toString in class Object

finish

public void finish(VectorAggregateExpression.AggregationBuffer aggregationBuffer,
                   boolean aborted)

Optional method to implement in VectorAggregateExpression instances. It's called for all aggregators from ProcessingMode, when VectorGroupByOperator is closed (before flush) or a write/flush happens. Calling this method properly before writing rows is the responsibility of VectorGroupByOperator.

Parameters:

aggregationBuffer -

aborted -

batchNeedsClone

public boolean batchNeedsClone()

This method should return true, when the particular VectorAggregateExpression must use a cloned batch while processing one, see Operator.batchNeedsClone() for further details.