Uses of Class
org.apache.sysds.lops.Lop

Packages that use Lop

Package	Description
org.apache.sysds.hops
org.apache.sysds.hops.codegen
org.apache.sysds.lops
org.apache.sysds.lops.compile
org.apache.sysds.lops.compile.linearization
org.apache.sysds.lops.rewrite
org.apache.sysds.parser
org.apache.sysds.runtime.instructions

Uses of Lop in org.apache.sysds.hops

Methods in org.apache.sysds.hops that return Lop

Modifier and Type	Method	Description
Lop	DnnOp.constructDnnLops(Types.ExecType et, List<Hop> inputs)
Lop	AggBinaryOp.constructLops()	NOTE: overestimated mem in case of transpose-identity matmult, but 3/2 at worst and existing mem estimate advantageous in terms of consistency hops/lops, and some special cases internally materialize the transpose for better cache locality
Lop	AggUnaryOp.constructLops()
Lop	BinaryOp.constructLops()
Lop	DataGenOp.constructLops()
Lop	DataOp.constructLops()
Lop	DnnOp.constructLops()
Lop	FunctionOp.constructLops()
abstract Lop	Hop.constructLops()
Lop	IndexingOp.constructLops()
Lop	LeftIndexingOp.constructLops()
Lop	LiteralOp.constructLops()
Lop	NaryOp.constructLops()	Construct the corresponding Lops for this Hop
Lop	ParameterizedBuiltinOp.constructLops()
Lop	QuaternaryOp.constructLops()
Lop	ReorgOp.constructLops()
Lop	TernaryOp.constructLops()
Lop	UnaryOp.constructLops()
static Lop	BinaryOp.constructSPAppendLop(Hop left, Hop right, Types.DataType dt, Types.ValueType vt, boolean cbind, Hop current)
static Lop	Hop.createOffsetLop(Hop hop, boolean repCols)
Lop	Hop.getLops()

Methods in org.apache.sysds.hops with parameters of type Lop

Modifier and Type	Method	Description
void	Hop.setLops(Lop lops)
void	Hop.updateLopFedOut(Lop lop)
void	Hop.updateLopFedOut(Lop lop, Types.ExecType execType, FEDInstruction.FederatedOutput fedOut)

Uses of Lop in org.apache.sysds.hops.codegen

Methods in org.apache.sysds.hops.codegen that return Lop

Modifier and Type	Method	Description
Lop	SpoofFusedOp.constructLops()

Uses of Lop in org.apache.sysds.lops

Subclasses of Lop in org.apache.sysds.lops

Modifier and Type	Class	Description
class	Append
class	AppendG	TODO Additional compiler enhancements: 1) Partial Shuffle Elimination - Any full or aligned blocks could be directly output from the mappers to the result index.
class	AppendGAlignedSP
class	AppendM
class	AppendR
class	Binary	Lop to perform binary operation.
class	BinaryM	Lop to perform binary operation.
class	BinaryScalar	Lop to perform binary scalar operations.
class	BinaryUAggChain
class	CentralMoment	Lop to perform cross product operation
class	Checkpoint	Lop for checkpoint operations.
class	Compression
class	CoVariance	Lop to compute covariance between two 1D matrices
class	CSVReBlock	Lop to convert CSV data into SystemDS data format
class	Ctable	Lop to perform ternary operation.
class	CumulativeOffsetBinary
class	CumulativePartialAggregate
class	Data	Lop to represent data objects.
class	DataGen	Defines a LOP that generates data.
class	DeCompression
class	DnnTransform
class	Federated
class	FunctionCallCP
class	GroupedAggregate	Lop to perform grouped aggregates
class	GroupedAggregateM	Lop to perform mr map-side grouped aggregates (restriction: sum, w/o weights, ngroups), groups broadcasted
class	LeftIndex
class	Local
class	MapMult
class	MapMultChain
class	MatMultCP
class	MMCJ	Lop to perform cross product operation
class	MMRJ	Lop to perform cross product operation
class	MMTSJ	Lop to perform transpose-identity operation (t(X)%*%X or X%*%t(X)), used to represent CP and MR instruction but in case of MR there is an additional Aggregate at the reducers.
class	MMZip	Lop to perform zip matrix multiplication
class	Nary	Lop to perform an operation on a variable number of operands.
class	ParameterizedBuiltin	Defines a LOP for functions.
class	PartialAggregate	Lop to perform a partial aggregation.
class	PickByCount
class	PMapMult
class	PMMJ
class	ReBlock	Lop to perform reblock operation
class	RightIndex
class	SortKeys
class	SpoofFused
class	Sql
class	Ternary	Lop to perform Sum of a matrix with another matrix multiplied by Scalar.
class	TernaryAggregate
class	Transform
class	UAggOuterChain	TODO Currently this lop only support the right hand side in distributed cache.
class	Unary	Lop to perform following operations: with one operand -- NOT(A), ABS(A), SQRT(A), LOG(A) with two operands where one of them is a scalar -- H=H*i, H=H*5, EXP(A,2), LOG(A,2)
class	UnaryCP
class	WeightedCrossEntropy
class	WeightedCrossEntropyR
class	WeightedDivMM
class	WeightedDivMMR
class	WeightedSigmoid
class	WeightedSigmoidR
class	WeightedSquaredLoss
class	WeightedSquaredLossR
class	WeightedUnaryMM
class	WeightedUnaryMMR

Methods in org.apache.sysds.lops that return Lop

Modifier and Type	Method	Description
Lop	AppendM.getBroadcastInput()
Lop	Binary.getBroadcastInput()
Lop	BinaryM.getBroadcastInput()
Lop	LeftIndex.getBroadcastInput()
Lop	Lop.getBroadcastInput()	Method to get the input to be broadcast.
Lop	MapMult.getBroadcastInput()
Lop	MapMultChain.getBroadcastInput()
Lop	PMMJ.getBroadcastInput()
Lop	Lop.getInput(int index)
Lop	ParameterizedBuiltin.getNamedInput(String name)
Lop	Data.getNamedInputLop(String name)
Lop	Data.getNamedInputLop(String name, String defaultVal)

Methods in org.apache.sysds.lops that return types with arguments of type Lop

Modifier and Type	Method	Description
ArrayList<Lop>	FunctionCallCP.getFunctionOutputs()
HashMap<String,Lop>	Data.getInputParams()	method to get inputParams
ArrayList<Lop>	Lop.getInputs()	Method to get input of Lops
static ArrayList<Lop>	OperatorOrderingUtils.getLopList(StatementBlock sb)
ArrayList<Lop>	Lop.getOutputs()	Method to get output of Lops

Methods in org.apache.sysds.lops with parameters of type Lop

Modifier and Type	Method	Description
void	Lop.addInput(Lop op)	Method to add input to Lop
void	Lop.addOutput(Lop op)	Method to add output to Lop
static int	OperatorOrderingUtils.collectGPURoots(Lop root, Map<Long,Integer> gpuOpCount, HashSet<Lop> gpuRoots)
static int	OperatorOrderingUtils.collectSparkRoots(Lop root, Map<Long,Integer> sparkOpCount, HashSet<Lop> sparkRoots)
static SortKeys	SortKeys.constructSortByValueLop(Lop input1, Lop input2, SortKeys.OperationTypes op, Types.DataType dt, Types.ValueType vt, Types.ExecType et, int numThreads)
static SortKeys	SortKeys.constructSortByValueLop(Lop input1, SortKeys.OperationTypes op, Types.DataType dt, Types.ValueType vt, Types.ExecType et, int numThreads)
void	Append.init(Lop input1, Lop input2, Lop input3, Types.DataType dt, Types.ValueType vt)
void	AppendG.init(Lop input1, Lop input2, Lop input3, Lop input4, Types.DataType dt, Types.ValueType vt, Types.ExecType et)
void	AppendGAlignedSP.init(Lop input1, Lop input2, Lop input3, Types.DataType dt, Types.ValueType vt)
void	AppendM.init(Lop input1, Lop input2, Lop input3, Types.DataType dt, Types.ValueType vt, Types.ExecType et)
void	AppendR.init(Lop input1, Lop input2, Types.DataType dt, Types.ValueType vt, Types.ExecType et)
static boolean	OperatorOrderingUtils.isCollectForBroadcast(Lop lop)
static boolean	OperatorOrderingUtils.isLopRoot(Lop lop)
static boolean	OperatorOrderingUtils.isPersistableSparkOp(Lop lop)
void	Lop.removeInput(Lop op)
void	Lop.removeOutput(Lop op)	Method to remove output from Lop
void	Lop.replaceInput(Lop oldInp, Lop newInp)	Method to replace an input to a Lop
static void	PartialAggregate.setDimensionsBasedOnDirection(Lop lop, long dim1, long dim2, long blen, Types.Direction dir)

Method parameters in org.apache.sysds.lops with type arguments of type Lop

Modifier and Type	Method	Description
void	Lop.addToDag(Dag<Lop> dag)	Method to recursively add LOPS to a DAG
static int	OperatorOrderingUtils.collectGPURoots(Lop root, Map<Long,Integer> gpuOpCount, HashSet<Lop> gpuRoots)
static int	OperatorOrderingUtils.collectSparkRoots(Lop root, Map<Long,Integer> sparkOpCount, HashSet<Lop> sparkRoots)
static void	OperatorOrderingUtils.markSharedSparkOps(HashSet<Lop> sparkRoots, Map<Long,Integer> operatorJobCount)
void	Lop.replaceAllInputs(ArrayList<Lop> newInputs)
void	Lop.replaceAllOutputs(ArrayList<Lop> newOutputs)
void	LopProperties.setLevel(ArrayList<Lop> inputs)
void	LopProperties.setProperties(ArrayList<Lop> inputs, Types.ExecType et)

Constructors in org.apache.sysds.lops with parameters of type Lop

Constructor	Description
Append(Lop input1, Lop input2, Lop input3, Types.DataType dt, Types.ValueType vt, boolean cbind, Types.ExecType et)
AppendG(Lop input1, Lop input2, Lop input3, Lop input4, Types.DataType dt, Types.ValueType vt, boolean cbind, Types.ExecType et)
AppendGAlignedSP(Lop input1, Lop input2, Lop input3, Types.DataType dt, Types.ValueType vt, boolean cbind)
AppendM(Lop input1, Lop input2, Lop input3, Types.DataType dt, Types.ValueType vt, boolean cbind, boolean partitioned, Types.ExecType et)
AppendR(Lop input1, Lop input2, Types.DataType dt, Types.ValueType vt, boolean cbind, Types.ExecType et)
Binary(Lop input1, Lop input2, Types.OpOp2 op, Types.DataType dt, Types.ValueType vt, Types.ExecType et)	Constructor to perform a binary operation.
Binary(Lop input1, Lop input2, Types.OpOp2 op, Types.DataType dt, Types.ValueType vt, Types.ExecType et, int k)
Binary(Lop input1, Lop input2, Types.OpOp2 op, Types.DataType dt, Types.ValueType vt, Types.ExecType et, int k, boolean inplace)
BinaryM(Lop input1, Lop input2, Types.OpOp2 op, Types.DataType dt, Types.ValueType vt, Types.ExecType et, boolean colVector)	Constructor to perform a binary operation.
BinaryScalar(Lop input1, Lop input2, Types.OpOp2 op, Types.DataType dt, Types.ValueType vt)	Constructor to perform a scalar operation
BinaryUAggChain(Lop input1, Types.OpOp2 bop, Types.AggOp uaop, Types.Direction uadir, Types.DataType dt, Types.ValueType vt, Types.ExecType et)	Constructor to setup a map mult chain without weights
CentralMoment(Lop input1, Lop input2, Types.DataType dt, Types.ValueType vt, int numThreads, Types.ExecType et)
CentralMoment(Lop input1, Lop input2, Lop input3, Types.DataType dt, Types.ValueType vt, int numThreads, Types.ExecType et)
Checkpoint(Lop input, Types.DataType dt, Types.ValueType vt, String level)
Checkpoint(Lop input, Types.DataType dt, Types.ValueType vt, String level, boolean isAsync)	TODO change string parameter storage.level to StorageLevel as soon as we can assume that Spark libraries are always available.
Compression(Lop input, Types.DataType dt, Types.ValueType vt, Types.ExecType et, int singletonLookupKey, int numThreads)
CoVariance(Lop input1, Lop input2, Types.DataType dt, Types.ValueType vt, int numThreads, Types.ExecType et)
CoVariance(Lop input1, Lop input2, Lop input3, Types.DataType dt, Types.ValueType vt, int numThreads, Types.ExecType et)
CSVReBlock(Lop input, int blen, Types.DataType dt, Types.ValueType vt, Types.ExecType et)
Ctable(Lop[] inputLops, Ctable.OperationTypes op, Types.DataType dt, Types.ValueType vt, boolean ignoreZeros, boolean outputEmptyBlocks, Types.ExecType et)
Ctable(Lop[] inputLops, Ctable.OperationTypes op, Types.DataType dt, Types.ValueType vt, Types.ExecType et)
CumulativeOffsetBinary(Lop data, Lop offsets, Types.DataType dt, Types.ValueType vt, double init, boolean broadcast, Types.AggOp op, Types.ExecType et)
CumulativeOffsetBinary(Lop data, Lop offsets, Types.DataType dt, Types.ValueType vt, Types.AggOp op, Types.ExecType et)
CumulativePartialAggregate(Lop input, Types.DataType dt, Types.ValueType vt, Types.AggOp op, Types.ExecType et)
Data(Types.OpOpData op, Lop input, HashMap<String,Lop> inputParametersLops, String name, String literal, Types.DataType dt, Types.ValueType vt, Types.FileFormat fmt)	Constructor to setup read or write LOP In case of write: input must be provided.
DeCompression(Lop input, Types.DataType dt, Types.ValueType vt, Types.ExecType et)
DnnTransform(Lop input, Types.OpOpDnn op, Types.DataType dt, Types.ValueType vt, Types.ExecType et, int k, double intermediateMemBudget)	Constructor when we have one input.
DnnTransform(Lop input1, Lop input2, Types.OpOpDnn op, Types.DataType dt, Types.ValueType vt, Types.ExecType et, int k)
DnnTransform(Lop input1, Lop input2, Lop input3, Types.OpOpDnn op, Types.DataType dt, Types.ValueType vt, Types.ExecType et, int k)
LeftIndex(Lop lhsInput, Lop rhsInput, Lop rowL, Lop rowU, Lop colL, Lop colU, Types.DataType dt, Types.ValueType vt, Types.ExecType et)
LeftIndex(Lop lhsInput, Lop rhsInput, Lop rowL, Lop rowU, Lop colL, Lop colU, Types.DataType dt, Types.ValueType vt, Types.ExecType et, LeftIndex.LixCacheType type)
Local(Lop input, Types.DataType dt, Types.ValueType vt)
MapMult(Lop input1, Lop input2, Types.DataType dt, Types.ValueType vt, boolean rightCache, boolean partitioned, boolean emptyBlocks, AggBinaryOp.SparkAggType aggtype)	Constructor to setup a partial Matrix-Vector Multiplication for Spark
MapMultChain(Lop input1, Lop input2, Types.DataType dt, Types.ValueType vt, Types.ExecType et)	Constructor to setup a map mult chain without weights
MapMultChain(Lop input1, Lop input2, Lop input3, MapMultChain.ChainType chain, Types.DataType dt, Types.ValueType vt, Types.ExecType et)	Constructor to setup a map mult chain with weights
MatMultCP(Lop input1, Lop input2, Types.DataType dt, Types.ValueType vt, Types.ExecType et)
MatMultCP(Lop input1, Lop input2, Types.DataType dt, Types.ValueType vt, Types.ExecType et, boolean isLeftTransposed, boolean isRightTransposed)
MatMultCP(Lop input1, Lop input2, Types.DataType dt, Types.ValueType vt, Types.ExecType et, int k)
MatMultCP(Lop input1, Lop input2, Types.DataType dt, Types.ValueType vt, Types.ExecType et, int k, boolean isLeftTransposed, boolean isRightTransposed)
MMCJ(Lop input1, Lop input2, Types.DataType dt, Types.ValueType vt, boolean outputEmptyBlocks, AggBinaryOp.SparkAggType aggtype, Types.ExecType et)
MMCJ(Lop input1, Lop input2, Types.DataType dt, Types.ValueType vt, MMCJ.MMCJType type, Types.ExecType et)	Constructor to perform a cross product operation.
MMRJ(Lop input1, Lop input2, Types.DataType dt, Types.ValueType vt, Types.ExecType et)	Constructor to perform a cross product operation.
MMTSJ(Lop input1, Types.DataType dt, Types.ValueType vt, Types.ExecType et, MMTSJ.MMTSJType type)
MMTSJ(Lop input1, Types.DataType dt, Types.ValueType vt, Types.ExecType et, MMTSJ.MMTSJType type, boolean multiPass)
MMTSJ(Lop input1, Types.DataType dt, Types.ValueType vt, Types.ExecType et, MMTSJ.MMTSJType type, boolean multiPass, int k)
MMZip(Lop input1, Lop input2, Types.DataType dt, Types.ValueType vt, boolean tRewrite, Types.ExecType et)
Nary(Types.OpOpN operationType, Types.DataType dt, Types.ValueType vt, Lop[] inputLops, Types.ExecType et)
PartialAggregate(Lop input, Types.AggOp op, Types.Direction direct, Types.DataType dt, Types.ValueType vt, Types.ExecType et, int k)
PartialAggregate(Lop input, Types.AggOp op, Types.Direction direct, Types.DataType dt, Types.ValueType vt, AggBinaryOp.SparkAggType aggtype, Types.ExecType et)
PickByCount(Lop input1, Lop input2, Types.DataType dt, Types.ValueType vt, PickByCount.OperationTypes op, Types.ExecType et)
PickByCount(Lop input1, Lop input2, Types.DataType dt, Types.ValueType vt, PickByCount.OperationTypes op, Types.ExecType et, boolean inMemoryInput)
PMapMult(Lop input1, Lop input2, Types.DataType dt, Types.ValueType vt)
PMMJ(Lop pminput, Lop rightinput, Lop nrow, Types.DataType dt, Types.ValueType vt, boolean partitioned, boolean emptyBlocks, Types.ExecType et)	Constructor to setup a Permutation Matrix Multiplication
ReBlock(Lop input, int blen, Types.DataType dt, Types.ValueType vt, boolean outputEmptyBlocks, Types.ExecType et)
RightIndex(Lop input, Lop rowL, Lop rowU, Lop colL, Lop colU, Types.DataType dt, Types.ValueType vt, Types.ExecType et)
RightIndex(Lop input, Lop rowL, Lop rowU, Lop colL, Lop colU, Types.DataType dt, Types.ValueType vt, Types.ExecType et, boolean forleft)
RightIndex(Lop input, Lop rowL, Lop rowU, Lop colL, Lop colU, Types.DataType dt, Types.ValueType vt, AggBinaryOp.SparkAggType aggtype, Types.ExecType et)
SortKeys(Lop input, boolean desc, SortKeys.OperationTypes op, Types.DataType dt, Types.ValueType vt, Types.ExecType et)
SortKeys(Lop input1, Lop input2, SortKeys.OperationTypes op, Types.DataType dt, Types.ValueType vt, Types.ExecType et, int numThreads)
SortKeys(Lop input, SortKeys.OperationTypes op, Types.DataType dt, Types.ValueType vt, Types.ExecType et, int numThreads)
Ternary(Types.OpOp3 op, Lop input1, Lop input2, Lop input3, Types.DataType dt, Types.ValueType vt, Types.ExecType et, int numThreads)
TernaryAggregate(Lop input1, Lop input2, Lop input3, Types.AggOp aggOp, Types.OpOp2 binOp, Types.Direction direction, Types.DataType dt, Types.ValueType vt, Types.ExecType et, int k)
Transform(Lop[] inputs, Types.ReOrgOp op, Types.DataType dt, Types.ValueType vt, boolean outputEmptyBlock, Types.ExecType et)
Transform(Lop[] inputs, Types.ReOrgOp op, Types.DataType dt, Types.ValueType vt, Types.ExecType et, boolean bSortIndInMem)
Transform(Lop[] inputs, Types.ReOrgOp op, Types.DataType dt, Types.ValueType vt, Types.ExecType et, boolean bSortIndInMem, int k)
Transform(Lop[] inputs, Types.ReOrgOp op, Types.DataType dt, Types.ValueType vt, Types.ExecType et, int k)
Transform(Lop input, Types.ReOrgOp op, Types.DataType dt, Types.ValueType vt, Types.ExecType et)
Transform(Lop input, Types.ReOrgOp op, Types.DataType dt, Types.ValueType vt, Types.ExecType et, boolean bSortIndInMem)
Transform(Lop input, Types.ReOrgOp op, Types.DataType dt, Types.ValueType vt, Types.ExecType et, int k)
UAggOuterChain(Lop input1, Lop input2, Types.AggOp uaop, Types.Direction uadir, Types.OpOp2 bop, Types.DataType dt, Types.ValueType vt, Types.ExecType et)	Constructor to setup a unaryagg outer chain
Unary(Lop input1, Types.OpOp1 op, Types.DataType dt, Types.ValueType vt, Types.ExecType et, int numThreads, boolean inplace)	Constructor to perform a unary operation with 1 input.
Unary(Lop input1, Lop input2, Types.OpOp1 op, Types.DataType dt, Types.ValueType vt, Types.ExecType et, int numThreads)	Constructor to perform a unary operation with 2 inputs
UnaryCP(Lop input, Types.OpOp1 op, Types.DataType dt, Types.ValueType vt)
UnaryCP(Lop input, Types.OpOp1 op, Types.DataType dt, Types.ValueType vt, int k)
UnaryCP(Lop input, Types.OpOp1 op, Types.DataType dt, Types.ValueType vt, Types.ExecType et)
UnaryCP(Lop input, Types.OpOp1 op, Types.DataType dt, Types.ValueType vt, Types.ExecType et, int k)	Constructor to perform a scalar operation
WeightedCrossEntropy(Lop input1, Lop input2, Lop input3, Lop input4, Types.DataType dt, Types.ValueType vt, WeightedCrossEntropy.WCeMMType wt, Types.ExecType et)
WeightedCrossEntropyR(Lop input1, Lop input2, Lop input3, Lop input4, Types.DataType dt, Types.ValueType vt, WeightedCrossEntropy.WCeMMType wt, boolean cacheU, boolean cacheV, Types.ExecType et)
WeightedDivMM(Lop input1, Lop input2, Lop input3, Lop input4, Types.DataType dt, Types.ValueType vt, WeightedDivMM.WDivMMType wt, Types.ExecType et)
WeightedDivMMR(Lop input1, Lop input2, Lop input3, Lop input4, Types.DataType dt, Types.ValueType vt, WeightedDivMM.WDivMMType wt, boolean cacheU, boolean cacheV, Types.ExecType et)
WeightedSigmoid(Lop input1, Lop input2, Lop input3, Types.DataType dt, Types.ValueType vt, WeightedSigmoid.WSigmoidType wt, Types.ExecType et)
WeightedSigmoidR(Lop input1, Lop input2, Lop input3, Types.DataType dt, Types.ValueType vt, WeightedSigmoid.WSigmoidType wt, boolean cacheU, boolean cacheV, Types.ExecType et)
WeightedSquaredLoss(Lop input1, Lop input2, Lop input3, Lop input4, Types.DataType dt, Types.ValueType vt, WeightedSquaredLoss.WeightsType wt, Types.ExecType et)
WeightedSquaredLossR(Lop input1, Lop input2, Lop input3, Lop input4, Types.DataType dt, Types.ValueType vt, WeightedSquaredLoss.WeightsType wt, boolean cacheU, boolean cacheV, Types.ExecType et)
WeightedUnaryMM(Lop input1, Lop input2, Lop input3, Types.DataType dt, Types.ValueType vt, WeightedUnaryMM.WUMMType wt, Types.OpOp1 op, Types.ExecType et)
WeightedUnaryMMR(Lop input1, Lop input2, Lop input3, Types.DataType dt, Types.ValueType vt, WeightedUnaryMM.WUMMType wt, Types.OpOp1 op, boolean cacheU, boolean cacheV, Types.ExecType et)

Constructor parameters in org.apache.sysds.lops with type arguments of type Lop

Constructor	Description
Data(Types.OpOpData op, Lop input, HashMap<String,Lop> inputParametersLops, String name, String literal, Types.DataType dt, Types.ValueType vt, Types.FileFormat fmt)	Constructor to setup read or write LOP In case of write: input must be provided.
DataGen(Types.OpOpDG op, DataIdentifier id, HashMap<String,Lop> inputParametersLops, String baseDir, Types.DataType dt, Types.ValueType vt, Types.ExecType et)	Creates a new Rand-LOP.
Federated(HashMap<String,Lop> inputLops, Types.DataType dataType, Types.ValueType valueType)
FunctionCallCP(ArrayList<Lop> inputs, String fnamespace, String fname, String[] inputNames, String[] outputNames, boolean opt, Types.ExecType et)
FunctionCallCP(ArrayList<Lop> inputs, String fnamespace, String fname, String[] inputNames, String[] outputNames, boolean opt, Types.ExecType et, int threads)
FunctionCallCP(ArrayList<Lop> inputs, String fnamespace, String fname, String[] inputNames, String[] outputNames, ArrayList<Hop> outputHops, boolean opt, Types.ExecType et)
FunctionCallCP(ArrayList<Lop> inputs, String fnamespace, String fname, String[] inputNames, String[] outputNames, ArrayList<Hop> outputHops, boolean opt, Types.ExecType et, int threads)
GroupedAggregate(HashMap<String,Lop> inputParameterLops, Types.DataType dt, Types.ValueType vt, Types.ExecType et)
GroupedAggregate(HashMap<String,Lop> inputParameterLops, Types.DataType dt, Types.ValueType vt, Types.ExecType et, boolean broadcastGroups)
GroupedAggregate(HashMap<String,Lop> inputParameterLops, Types.DataType dt, Types.ValueType vt, Types.ExecType et, int k)
GroupedAggregateM(HashMap<String,Lop> inputParameterLops, Types.DataType dt, Types.ValueType vt, boolean partitioned, Types.ExecType et)
ParameterizedBuiltin(HashMap<String,Lop> paramLops, Types.ParamBuiltinOp op, Types.DataType dt, Types.ValueType vt, Types.ExecType et)
ParameterizedBuiltin(HashMap<String,Lop> paramLops, Types.ParamBuiltinOp op, Types.DataType dt, Types.ValueType vt, Types.ExecType et, boolean bRmEmptyBC)
ParameterizedBuiltin(HashMap<String,Lop> paramLops, Types.ParamBuiltinOp op, Types.DataType dt, Types.ValueType vt, Types.ExecType et, int k)
SpoofFused(ArrayList<Lop> inputs, Types.DataType dt, Types.ValueType vt, Class<?> cla, SpoofCompiler.GeneratorAPI api, String genVarName, int k, Types.ExecType etype)
Sql(HashMap<String,Lop> inputParametersLops, Types.DataType dt, Types.ValueType vt)

Uses of Lop in org.apache.sysds.lops.compile

Classes in org.apache.sysds.lops.compile with type parameters of type Lop

Modifier and Type	Class	Description
class	Dag<N extends Lop>	Class to maintain a DAG of lops and compile it into runtime instructions, incl piggybacking into jobs.
class	LopComparator<N extends Lop>	Comparator class used in sorting the LopDAG in topological order.

Methods in org.apache.sysds.lops.compile with parameters of type Lop

Modifier and Type	Method	Description
boolean	Dag.addNode(Lop node)	Method to add a node to the DAG.

Uses of Lop in org.apache.sysds.lops.compile.linearization

Methods in org.apache.sysds.lops.compile.linearization that return types with arguments of type Lop

Modifier and Type	Method	Description
abstract List<Lop>	IDagLinearizer.linearize(List<Lop> v)	Linearized a DAG of lops into a sequence of lops that preserves all data dependencies.
List<Lop>	LinearizerBreadthFirst.linearize(List<Lop> v)
List<Lop>	LinearizerCostBased.linearize(List<Lop> v)
List<Lop>	LinearizerDepthFirst.linearize(List<Lop> v)
List<Lop>	LinearizerMaxParallelism.linearize(List<Lop> v)
List<Lop>	LinearizerMinIntermediates.linearize(List<Lop> v)
List<Lop>	LinearizerPipelineAware.linearize(List<Lop> v)	Sort lops depth-first while assigning the nodes to pipelines

Method parameters in org.apache.sysds.lops.compile.linearization with type arguments of type Lop

Modifier and Type	Method	Description
abstract List<Lop>	IDagLinearizer.linearize(List<Lop> v)	Linearized a DAG of lops into a sequence of lops that preserves all data dependencies.
List<Lop>	LinearizerBreadthFirst.linearize(List<Lop> v)
List<Lop>	LinearizerCostBased.linearize(List<Lop> v)
List<Lop>	LinearizerDepthFirst.linearize(List<Lop> v)
List<Lop>	LinearizerMaxParallelism.linearize(List<Lop> v)
List<Lop>	LinearizerMinIntermediates.linearize(List<Lop> v)
List<Lop>	LinearizerPipelineAware.linearize(List<Lop> v)	Sort lops depth-first while assigning the nodes to pipelines

Uses of Lop in org.apache.sysds.lops.rewrite

Methods in org.apache.sysds.lops.rewrite that return types with arguments of type Lop

Modifier and Type	Method	Description
ArrayList<Lop>	LopRewriter.rewriteLopDAG(StatementBlock sb, ArrayList<Lop> lops)

Method parameters in org.apache.sysds.lops.rewrite with type arguments of type Lop

Modifier and Type	Method	Description
ArrayList<Lop>	LopRewriter.rewriteLopDAG(StatementBlock sb, ArrayList<Lop> lops)

Uses of Lop in org.apache.sysds.parser

Methods in org.apache.sysds.parser that return Lop

Modifier and Type	Method	Description
Lop	ForStatementBlock.getFromLops()
Lop	ForStatementBlock.getIncrementLops()
Lop	IfStatementBlock.getPredicateLops()
Lop	WhileStatementBlock.getPredicateLops()
Lop	ForStatementBlock.getToLops()

Methods in org.apache.sysds.parser that return types with arguments of type Lop

Modifier and Type	Method	Description
ArrayList<Lop>	StatementBlock.getLops()

Methods in org.apache.sysds.parser with parameters of type Lop

Modifier and Type	Method	Description
void	StatementBlock.setCheckpointPosition(Lop input, List<Lop> outputs)
void	ForStatementBlock.setFromLops(Lop lops)
void	ForStatementBlock.setIncrementLops(Lop lops)
void	IfStatementBlock.setPredicateLops(Lop predicateLops)
void	WhileStatementBlock.setPredicateLops(Lop predicateLops)
void	ForStatementBlock.setToLops(Lop lops)

Method parameters in org.apache.sysds.parser with type arguments of type Lop

Modifier and Type	Method	Description
void	StatementBlock.setCheckpointPosition(Lop input, List<Lop> outputs)
void	StatementBlock.setLops(ArrayList<Lop> lops)

Uses of Lop in org.apache.sysds.runtime.instructions

Methods in org.apache.sysds.runtime.instructions with parameters of type Lop

Modifier and Type	Method	Description
void	Instruction.setLocation(Lop lop)