Package | Description |
---|---|
org.apache.sysds.runtime.controlprogram.caching | |
org.apache.sysds.runtime.data | |
org.apache.sysds.runtime.matrix.data | |
org.apache.sysds.runtime.util |
Modifier and Type | Method and Description |
---|---|
long |
CacheDataInput.readSparseRows(int rlen,
long nnz,
SparseBlock rows) |
void |
CacheDataOutput.writeSparseRows(int rlen,
SparseBlock rows) |
Modifier and Type | Class and Description |
---|---|
class |
SparseBlockCOO
SparseBlock implementation that realizes a traditional 'coordinate matrix'
representation, where the entire sparse block is stored as triples in three arrays:
row indexes, column indexes, and values, where row indexes and colunm indexes are
sorted in order to allow binary search.
|
class |
SparseBlockCSR
SparseBlock implementation that realizes a traditional 'compressed sparse row'
representation, where the entire sparse block is stored as three arrays: ptr
of length rlen+1 to store offsets per row, and indexes/values of length nnz
to store column indexes and values of non-zero entries.
|
class |
SparseBlockMCSR
SparseBlock implementation that realizes a 'modified compressed sparse row'
representation, where each compressed row is stored as a separate SparseRow
object which provides flexibility for unsorted row appends without the need
for global reshifting of values/indexes but it incurs additional memory
overhead per row for object/array headers per row which also slows down
memory-bound operations due to higher memory bandwidth requirements.
|
Modifier and Type | Method and Description |
---|---|
static SparseBlock |
SparseBlockFactory.copySparseBlock(SparseBlock.Type type,
SparseBlock sblock,
boolean forceCopy) |
static SparseBlock |
SparseBlockFactory.createSparseBlock(int rlen) |
static SparseBlock |
SparseBlockFactory.createSparseBlock(SparseBlock.Type type,
int rlen) |
static SparseBlock |
SparseBlockFactory.createSparseBlock(SparseBlock.Type type,
SparseRow row) |
SparseBlock |
BasicTensorBlock.getSparseBlock() |
Modifier and Type | Method and Description |
---|---|
static SparseBlock |
SparseBlockFactory.copySparseBlock(SparseBlock.Type type,
SparseBlock sblock,
boolean forceCopy) |
static SparseBlock.Type |
SparseBlockFactory.getSparseBlockType(SparseBlock sblock) |
boolean |
SparseBlock.isAligned(int r,
SparseBlock that)
Indicates if all non-zero values of row r are aligned with
the same row of the given second sparse block instance, which
can be exploited for more efficient operations.
|
boolean |
SparseBlock.isAligned(SparseBlock that)
Indicates if all non-zero values are aligned with the given
second sparse block instance, which can be exploited for
more efficient operations.
|
static boolean |
SparseBlockFactory.isSparseBlockType(SparseBlock sblock,
SparseBlock.Type type) |
void |
SparseBlockCSR.setIndexRange(int rl,
int ru,
int cl,
int cu,
SparseBlock sb)
Inserts a sparse block into the row and column range [rl,ru) and [cl,cu).
|
Constructor and Description |
---|
SparseBlockCOO(SparseBlock sblock)
Copy constructor sparse block abstraction.
|
SparseBlockCSR(SparseBlock sblock)
Copy constructor sparse block abstraction.
|
SparseBlockMCSR(SparseBlock sblock)
Copy constructor sparse block abstraction.
|
Modifier and Type | Method and Description |
---|---|
SparseBlock |
MatrixBlock.getSparseBlock() |
Modifier and Type | Method and Description |
---|---|
void |
MatrixBlock.appendRowToSparse(SparseBlock dest,
MatrixBlock src,
int i,
int rowoffset,
int coloffset,
boolean deep) |
long |
MatrixBlockDataInput.readSparseRows(int rlen,
long nnz,
SparseBlock rows)
Reads the sparse rows array from the data input into a sparse block
and returns the number of non-zeros.
|
void |
MatrixBlock.setSparseBlock(SparseBlock sblock) |
void |
MatrixBlockDataOutput.writeSparseRows(int rlen,
SparseBlock rows)
Writes the sparse rows array of a sparse block to the data output.
|
Constructor and Description |
---|
MatrixBlock(int rl,
int cl,
long nnz,
SparseBlock sblock)
Constructs a sparse
MatrixBlock with a given instance of a SparseBlock |
Modifier and Type | Method and Description |
---|---|
static long |
UtilFunctions.computeNnz(SparseBlock a,
int[] aix,
int ai,
int alen) |
long |
ByteBufferDataInput.readSparseRows(int rlen,
long nnz,
SparseBlock rows) |
long |
FastBufferedDataInputStream.readSparseRows(int rlen,
long nnz,
SparseBlock rows) |
void |
FastBufferedDataOutputStream.writeSparseRows(int rlen,
SparseBlock rows) |
Copyright © 2020 The Apache Software Foundation. All rights reserved.