Dew.Math Unit

Classes

NameDescription
GridServicesIntroduces rutines for copying vector or matrix values to and from System.Windows.Forms.DataGridView control. These routines are only for backward compatibility with older Dew Math version. To display vector or matrix in DataGridView, use new Dew.Math.Controls.MtxGridView control.
MtxVecBytesConverterGeneric converter for types implementing Dew.Math.IMtxVecSerializable. Serializes to/from System.Byte[] and shows a concise, read-only preview string in the PropertyGrid. Uses ComplexValues / IntegerValues for known Dew types.
MtxVecJsonSerializerCustom JsonConverter for TMtxVecBase and its known descendants (TMtx, TVec, TMtxInt, TVecInt). Stores instances as Base64-encoded byte[] using TMtxVecBase.GetObjectToByteArray/SetObjectFromByteArray. AOT/linker safe for Android/iOS via explicit factory map; cross-platform on Windows/Linux/macOS.
MtxVecPreviewConverterProvides a concise string preview for MtxVec objects (vector/matrix, real/integer). The preview is clipped to 50 characters for compact display in property editors.
TAbstractMtxVecCacheAbstract class for object cache.
TAbstractMtxVecIntIntermediate class to allow insertion of helper classes for TMtxVecInt.
TAMDDeviceInfoContains additional device information for AMD devices.
TBoolMatrixValue
TBoolValue
TBoolVectorValue
TComplexValue
TCustomValue
TDoubleValue
TEigBalancingClass used to store matrix balancing information and results returned by TMtx.Eig.. functions.
TEventThread
TExprBoolConstant
TExprBoolMatrixConstant
TExprBoolMatrixVariable
TExprBoolVariable
TExprBoolVectorConstant
TExprBoolVectorVariable
TExprComma
TExprComplexConstant
TExprComplexVariable
TExprConstant
TExprContextExpression context, holding expression list and variables with type and the data they hold.
TExprCustomValueVariable
TExprDoubleConstant
TExprDoubleVariable
TExprExtRangeConstant
TExprFunction
TExprGridVariableCustom variable object type.
TExprIntegerConstant
TExprIntegerVariable
TExprIntMatrixConstant
TExprIntMatrixVariable
TExprIntVectorConstant
TExprIntVectorVariable
TExprLeftBracket
TExprLeftSquareBracket
TExprList
TExprMatrixConstant
TExprMatrixVariable
TExprOperator
TExprRangeConstant
TExprRangeVariable
TExprRec
TExprResultIndexer
TExprRightBracket
TExprRightSquareBracket
TExprSemicolon
TExprStringConstant
TExprStringListConstant
TExprStringVariable
TExprVariable
TExprVaryingFunction
TExprVectorConstant
TExprVectorVariable
TExprWord
TExprWordList
TExprWorkspaceStores workspace for expression parser.
TFairSemaphoreSectionCritical section with semaphore capability.
TFifoCriticalSectionEnables threads in a multithreaded application a single-thread only exclusive execution of a part of the code.
TFPInfoContains device information specific to floating point math.
TIntegersList
TIntegerValue
TInternalProgressThreadInternal background worker for CLR targets.
TIntMatrixValue
TIntVectorValue
TIterativeSettings
TJobThread
TLoopThreadThread type used by TMtxForLoop.
TMatrixValue
TMtxTwo dimensional array - matrix.
TMtxByteClass for 8bit two dimensional integer arrays.
TMtxCache
TMtxCollection
TMtxCollectionItem
TMtxComponent
TMtxComponentListManages a list of TMtxComponents.
TMtxDoubleJaggedConverterConverts Dew.Math.TMtx to/from System.Double[][] for designer serialization,
TMtxExpressionExpression evaluator.
TMtxForLoopThe object implements everything neccessary to simplify threading of loops.
TMtxFunctionEvaluatorFunction evaluator.
TMtxIntTwo dimensional array - matrix.
TMtxIntCacheImplementation of object cache for TVecInt objects.
TMtxIntJaggedConverterConverts Dew.Math.TMtxInt to/from System.Int32[][] and displays Dew.Math.TMtxInt.IntegerValues.
TMtxListList of TMtx objects.
TMtxLPInterfaces Linear Programming algorithms.
TMtxMulJitCache
TMtxOptimizationInterfaces the optimization routines.
TMtxProgressRuntimeRuntime for background execution and progress reporting.
TMtxSmallIntClass for 16bit two dimensional integer arrays.
TMtxVecControllerControlls the object cache and threading logic. Provides multi-threading functions for MtxVec.
TMtxVecIntHelperInjects additional methods in to TMtxVecInt.
TMtxVecPoolItem
TMtxVecReportThe report printer.
TMtxVecSample
TNamedVariableNamed variables collection item.
TNVDeviceInfoContains additional device information for NVIDIA devices.
TOpenCLBase
TOpenCLCache
TOpenCLCommandQueueEvery Open CL function that we want to compute needs to be submitted to the command queue.
TOpenCLCommandQueueListStores a list of Open CL queues.
TOpenCLDeviceStores data and information associated with the Open CL computing device.
TOpenCLKernelOpen CL Kernels are added automatically to Dew.Math.TOpenCLDevice.Kernels list when a program is loaded.
TOpenCLKernelListStores a list of Open CL Kernels.
TOpenCLMatrix
TOpenCLPlatform
TOpenCLPlatformListHolds the list of all available platforms on the computer.
TOpenCLProgramProvides method to load Open CL programs either from resource, source or files.
TOpenCLProgramListStores a list of Open CL programs.
TOpenCLValueStores a single real or complex value in GPU memory.
TOpenCLVector
TOptControlOptional object class type to be passed to optimization routines.
TPardisoSettingsClass encapsulates Pardiso sparse solver specific settings.
TPardisoWork
TPiecePolyManages piece-wise polynomials.
TRandomGeneratorRandom generator.
TRangeValue
TRCustomData
TReferenceList
TReferenceListItem
TRngStreamRandom stream class.
TSingleExpression
TSmallMatrixMultiplyProvides interface for multiplying small matrices by using the lapack's dgemm api standard.)
TSparseMtxEncapsulates sparse matrix properties and methods/operations.
TStringListValue
TStringValue
TTaucsSettingsClass encapsulates Taucs sparse solver specific settings.
TTaucsWork
TValueRec
TVarByNameIndexer
TVariableCollectionCollection of Dew.Math.TNamedVariables.
TVecOne dimensional array - vector.
TVecByteClass for 8bit one dimensional integer arrays.
TVecCache
TVecDoubleArrayConverterConverts Dew.Math.TVec to/from System.Double[] for designer serialization,
TVecIntClass for one dimensional integer arrays.
TVecIntArrayConverterConverts Dew.Math.TVecInt to/from System.Int32[] and displays Dew.Math.TVecInt.IntegerValues.
TVecIntCacheImplementation of object cache for TVecInt objects.
TVecIntListList of TVecInt objects.
TVecListList of TVec objects.
TVecSmallIntClass for 16bit one dimensional integer arrays.
TVectorValue
TWordList

Interfaces

NameDescription
IMtxVecSerializableBinary serialization interface for Dew Math vector and matrix types. Provides a stable binary format compatible with Delphi.

Records

NameDescription
clMatrixOpen CL Matrix class with overloaded operators.
clValueOpen CL single value class with overloaded operators.
clVectorOpen CL vector class with overloaded operators.
MatrixMatrix class for operator overloading in D2006 and later.
MatrixIntInteger vector class for operator overloading (Delphi 2006 and later).
TBytec
TCardinalc
TCplxDefault complex number type used by Math387.
TDCplxComplex number type for double real and complex part.
TICplxComplex number type for integer real and complex part.
TInt64c
TIntegerc
TIntegerCplxComplex number type for integer real and complex part.
TLocalCPUIDMapped from Delphi System Unit
TMKL_Matrix_descr
TMtxRecord
TPoolItemRecord
TResHeader
TSCplxDefault complex number type used by Math387.
TShortintc
TSimpleCriticalSectionEnables threads in a multithreaded application a single-thread only exclusive execution of a part of the code.
TSmallintc
TSmallIntCplxComplex number type for integer real and complex part.
TSubRangeItem
TVecRecord
TWordc
VectorVector class for operator overloading (Delphi 2006 and later).
VectorIntInteger vector class for operator overloading (Delphi 2006 and later).

Enumerations

NameDescription
TBalanceTypeDefines the method, used to balance the matrix.
TBasicRandomGeneratorDefines available random generator algorithms.
TclFloatPrecisionSpecifies the precision to be used for Open CL computation.
TCompareComparison type for the Equal function.
TConditionNumberDefines the norm, used in calculating the condition number parameter.
TConvergenceCheckThe criteria for convergence check.
TCorrNormDefines type of correlation norm used for Cross-Correlation and Auto-Correlation.
TCRWaitModeThe waiting method used by the sorted critical section.
TDistributionAll supported probability distributions.
TEigConditionSpecifies if and which condition numbers for eigenvalues are to be computed.
TEigGenTypeDefines the type of the symmetric-definite generalized eigenvalue problem.
TEndiannessStorage Endianness.
TEquidistantSampleDefines the type of pixel downsampling.
TEventStateInternal event state used by TMtxForLoop.
TExprFlowControl
TExprFlowKeyword
TFFTStorageFormatType used to define the storage format for FFT operations.
TFixedTextAlignSpecifies alignment for fixed width font number to string conversion.
TGaussRandMethodGaussian (normal) distribution random generator options.
THilbertMethodFast Hilbert Transform computation method.
TIndexTypeType used to define indexation type.
TIntegrateAlgoDefines integration algorithm.
TInterpolationTypeDefines the type of 1D interpolation.
TIntPrecisionDefault storage precision for integer vector and matrix values.
TIntStopReasonDefines numerical integration algorithm exit reason.
TIterativeMethodThe method for iterative solver.
TIterativePreconditionerIterative preconditioner options.
TIterativeSolverMethod
TJITedMulDefines the type of operation on matrix.
TLocalMemType
TLPAlgorithmLinear programming algorithm.
TLPSolutionLP system solution.
TMKL_Compact_Pack
TMKL_Diag
TMKL_Jit_Status
TMKL_Layout
TMKL_Side
TMKL_Sparse_diag_typeApplies to triangular matrices only ( SPARSE_MATRIX_TYPE_SYMMETRIC, SPARSE_MATRIX_TYPE_HERMITIAN, SPARSE_MATRIX_TYPE_TRIANGULAR )
TMKL_Sparse_fill_modeApplies to triangular matrices only ( SPARSE_MATRIX_TYPE_SYMMETRIC, SPARSE_MATRIX_TYPE_HERMITIAN, SPARSE_MATRIX_TYPE_TRIANGULAR )
TMKL_Sparse_index_base
TMKL_Sparse_layoutApplicable for Level 3 operations with dense matrices; describes storage scheme for dense matrix (row major or column major)
TMKL_Sparse_matrix_type
TMKL_Sparse_memory_usagememory optimization hints from user: describe how much memory could be used on optimization stage
TMKL_Sparse_operation
TMKL_Sparse_request
TMKL_Sparse_sor_typeApplies to SOR interface; define type of (S)SOR operation to perform
TMKL_Sparse_status
TMKL_Transpose
TMKL_Uplo
TMKL_Verbose_modeIf verbose mode activated, handle should collect and report profiling / optimization info
TMklThreadingModeSets threading mode used by MtxVec.
TMtxCompareOp
TMtxFloatPrecisionSpecifies MtxVec processing precision.
TMtxOperationDefines the type of operation on matrix.
TMtxProgressEventDefines the thread event type.
TMtxTypeDefines type of matrix.
TMtxVecType
TNumericGradMethodNumerical gradient approximation methods.
TOptMethodOptimization methods.
TOptStopReasonStop reason for the main loop in optimization.
TOutofCoreModeOut-of core sparse matrix solver.
TParallelFactorizationParallel factorization control.
TParallelForwardBackwardParallel solve control.
TPardisoIterativeIterative solution refinment for Pardiso solver.
TPardisoReorderingThe reordering method for Pardiso solver.
TPixelDownSampleDefines type of size reduction used in PixelDownSample method.
TPoissonRandMethodPossion distribution random generator options.
TPrecisionStorage precision.
TQuadMethodDefines weight type for Gauss quadrature integration algorithm.
TReportLevelThe detail of the report generated by Dew.Math.TSparseMtx.Solve method.
TRngStreamIndependenceDefines the method used to get random number series independence.
TRoundingRounding algorithm.
TSignDefines the shape of the Sylvester equation.
TSimpleType
TSparseFormatDefines sparse matrix file format.
TSparsePatternDefines how to solve the matrix using the Solve method.
TSparseSolverDefines which method will be used to solve system of equations.
TSparseSystemDefines the system to be solved by the Dew.Math.TSparseMtx.Solve method.
TSubRangeOp
TTaucsFactorizationIterative solution refinment for Taucs.
TTaucsIterativeIterative solution refinment for Taucs.
TTaucsReorderingDefines the reordering method used by Taucs.
TThreadAffinityModeDefines how the threads can be assigned to specific CPU cores.
TThreadingModeSpecifies threading mode for TMtxForLoop.
TTriangleFormDefines the matrix triangular type.
TVectorFormDefines the vector form in eigenvector calculation.

Delegates

NameDescription
Tcgemm_jit_kernel_t
Tcgemm_jit_kernel_tp
TComplexFunc
TComputeEventCompute event signature.
Tdgemm_jit_kernel_t
Tdgemm_jit_kernel_tp
TDoubleFunc
TExceptionEventException event signature.
TExprFuncFunction type for custom function definition.
TExprFuncObj
TExprGridGetProc
TExprGridSetProc
TExprTest
TForLoopEventEvent type to implement threaded for-loop within a method of object.
TForLoopRangeEventEvent type to implement threaded for-loop within a function.
Tgemm_jit_kernel_tp
TGradDefines the procedure for calculating the gradient of a real function.
TGradHessDefines the procedure for calculating the gradient and Hessian matrix of a real function.
TIterativeStopTestThe event type called when user defined stop criteria for iterative sparse solver is requested.
TJacobianFunctionDefines procedure for calculating the Jacobian matrix.
TMKLTRFunction
TMtxComponentAddEventEvent for TMtxComponentList.
TMtxComponentDeleteEventEvent for TMtxComponentList.
TMtxFunctionDefines the matrix function.
TMtxNotifyEventBase notify event.
TOnClExecuteEvent
TOnPreprocessorEvent type for the preprocessor.
TProgressUpdateEventProgress update event signature.
TRealFunctionDefines real scalar function of several variables.
TSComplexFunc
TSelectOneFunctionCFunction type with two parameters.
TSelectOneFunctionSingleCFunction type with two parameters.
TSelectThreeFunctionFunction type with two parameters.
TSelectThreeFunctionSingleFunction type with two parameters.
TSelectTwoFunctionFunction type with two parameters.
TSelectTwoFunctionCFunction type with two parameters.
TSelectTwoFunctionSingleFunction type with two parameters.
TSelectTwoFunctionSingleCFunction type with two parameters.
Tsgemm_jit_kernel_t
Tsgemm_jit_kernel_tp
TSingleFunc
TVectorFunctionDefines vector function of several variables.
TWhileLoopEventEvent type to implement threaded while-loop within a method of object.
TWordPopulate
Tzgemm_jit_kernel_t
Tzgemm_jit_kernel_tp

Exceptions

NameDescription
EMtxParseErrorParser exception error class.
EMtxVecExceptionBase exception class for MtxVec library.
EMtxVecInvalidArgumentException class indicates invalid arguments error.
EMtxVecRangeErrorException class indicates out of range error.
ENoOperationException type raised when operation is not supported.
EOutOfMemoryException class indicates out of memory error.