Package org.apache.sysds.runtime.compress.colgroup

Class AMorphingMMColGroup

java.lang.Object

org.apache.sysds.runtime.compress.colgroup.AColGroup

org.apache.sysds.runtime.compress.colgroup.AColGroupCompressed

org.apache.sysds.runtime.compress.colgroup.ADictBasedColGroup

org.apache.sysds.runtime.compress.colgroup.AColGroupValue

org.apache.sysds.runtime.compress.colgroup.AMorphingMMColGroup

All Implemented Interfaces:

Serializable, IContainADictionary

Direct Known Subclasses:

ASDC, ColGroupDDCFOR

public abstract class AMorphingMMColGroup
extends AColGroupValue

Abstract class for column group types that do not perform matrix Multiplication, and decompression for performance reasons but instead transforms into another type of column group type to perform that operation.

See Also:

• Serialized Form

Nested Class Summary

Nested classes/interfaces inherited from class org.apache.sysds.runtime.compress.colgroup.AColGroup

AColGroup.CompressionType

Method Summary

Modifier and Type	Method	Description
void	decompressToSparseBlockTransposed(SparseBlockMCSR db, int nColOut)	Decompress the column group to the sparse transposed block.
abstract AColGroup	extractCommon(double[] constV)	extract common value from group and return non morphing group
abstract double[]	getCommon()	Get common vector, note this should not materialize anything but simply point to things that are already allocated.
final void	leftMultByAColGroup(AColGroup lhs, MatrixBlock result, int nRows)	Left side matrix multiplication with a column group that is transposed.
final void	leftMultByMatrixNoPreAgg(MatrixBlock matrix, MatrixBlock result, int rl, int ru, int cl, int cu)	Left multiply with this column group.
final void	tsmmAColGroup(AColGroup other, MatrixBlock result)	Matrix multiply with this other column group, but: 1.

Methods inherited from class org.apache.sysds.runtime.compress.colgroup.AColGroupValue

centralMoment, clear, computeColSums, estimateInMemorySize, getCounts, getNumberNonZeros, getNumValues, replace, rexpandCols, toString

Methods inherited from class org.apache.sysds.runtime.compress.colgroup.ADictBasedColGroup

copyAndSet, copyAndSet, decompressToDenseBlock, decompressToDenseBlockTransposed, decompressToSparseBlock, getDictionary, getExactSizeOnDisk, getSparsity, reduceCols, rightMultByMatrix, write

Methods inherited from class org.apache.sysds.runtime.compress.colgroup.AColGroupCompressed

getMax, getMin, getSum, isEmpty, preAggRows, sameIndexStructure, sameIndexStructure, tsmm, unaryAggregateOperations, unaryAggregateOperations

Methods inherited from class org.apache.sysds.runtime.compress.colgroup.AColGroup

addVector, append, appendN, binaryRowOpLeft, binaryRowOpRight, colSum, combine, combineWithSameIndex, combineWithSameIndex, containsValue, decompressToDenseBlock, decompressToSparseBlock, get, getColIndices, getCompressionInfo, getCompressionScheme, getCompType, getCost, getEncoding, getIdx, getNumCols, morph, recompress, removeEmptyCols, removeEmptyRows, rightDecompressingMult, rightMultByMatrix, scalarOperation, selectionMultiply, shiftColIndices, sliceColumn, sliceColumns, sliceRows, sort, sortColumnIndexes, splitReshape, splitReshapePushDown, unaryOperation

Methods inherited from class java.lang.Object

equals, getClass, hashCode, notify, notifyAll, wait, wait, wait

Method Details

decompressToSparseBlockTransposed

public void decompressToSparseBlockTransposed(SparseBlockMCSR db,
 int nColOut)

Description copied from class: AColGroup

Decompress the column group to the sparse transposed block. Note that the column groups would only need to decompress into specific sub rows of the Sparse block

Overrides:

decompressToSparseBlockTransposed in class ADictBasedColGroup

Parameters:

db - Sparse target block

nColOut - The number of columns in the sb.

leftMultByMatrixNoPreAgg

public final void leftMultByMatrixNoPreAgg(MatrixBlock matrix,
 MatrixBlock result,
 int rl,
 int ru,
 int cl,
 int cu)

Description copied from class: AColGroup

Left multiply with this column group.

Specified by:

leftMultByMatrixNoPreAgg in class AColGroup

Parameters:

matrix - The matrix to multiply with on the left

result - The result to output the values into, always dense for the purpose of the column groups parallelizing

rl - The row to begin the multiplication from on the lhs matrix

ru - The row to end the multiplication at on the lhs matrix

cl - The column to begin the multiplication from on the lhs matrix

cu - The column to end the multiplication at on the lhs matrix

leftMultByAColGroup

public final void leftMultByAColGroup(AColGroup lhs,
 MatrixBlock result,
 int nRows)

Description copied from class: AColGroup

Left side matrix multiplication with a column group that is transposed.

Specified by:

leftMultByAColGroup in class AColGroup

Parameters:

lhs - The left hand side Column group to multiply with, the left hand side should be considered transposed. Also it should be guaranteed that this column group is not empty.

result - The result matrix to insert the result of the multiplication into

nRows - Number of rows in the lhs colGroup

tsmmAColGroup

public final void tsmmAColGroup(AColGroup other,
 MatrixBlock result)

Description copied from class: AColGroup

Matrix multiply with this other column group, but: 1. Only output upper triangle values. 2. Multiply both ways with "this" being on the left and on the right. It should be guaranteed that the input is not the same as the caller of the method. The second step is achievable by treating the initial multiplied matrix, and adding its values to the correct locations in the output.

Specified by:

tsmmAColGroup in class AColGroup

Parameters:

other - The other Column group to multiply with

result - The result matrix to put the results into

extractCommon

public abstract AColGroup extractCommon(double[] constV)

extract common value from group and return non morphing group

Parameters:

constV - a vector to contain all values, note length = nCols in total matrix.

Returns:

A non morphing column group with decompression instructions.

getCommon

public abstract double[] getCommon()

Get common vector, note this should not materialize anything but simply point to things that are already allocated.

Returns:

the common double vector