|
Tpetra parallel linear algebra
Version of the Day
|
|
Tpetra parallel linear algebra
Version of the Day
|
A distributed dense vector. More...
#include <Tpetra_Vector_decl.hpp>
Public Types | |
Typedefs to facilitate template metaprogramming | |
| typedef Scalar | scalar_type |
| This class' first template parameter; the type of each entry in the Vector. More... | |
| typedef base_type::impl_scalar_type | impl_scalar_type |
The type used internally in place of Scalar. More... | |
| typedef LocalOrdinal | local_ordinal_type |
| This class' second template parameter; the type of local indices. More... | |
| typedef GlobalOrdinal | global_ordinal_type |
| This class' third template parameter; the type of global indices. More... | |
| typedef Node::device_type | device_type |
| The Kokkos device type. More... | |
| typedef Node | node_type |
| The Kokkos Node type. More... | |
| typedef base_type::dot_type | dot_type |
| Type of an inner ("dot") product result. More... | |
| typedef base_type::mag_type | mag_type |
| Type of a norm result. More... | |
| typedef base_type::dual_view_type | dual_view_type |
| Kokkos::DualView specialization used by this class. More... | |
| typedef base_type::map_type | map_type |
| The type of the Map specialization used by this class. More... | |
Typedefs to facilitate template metaprogramming. | |
| using | execution_space = typename device_type::execution_space |
| Type of the (new) Kokkos execution space. More... | |
Typedefs | |
| using | packet_type = typename::Kokkos::Details::ArithTraits< Scalar >::val_type |
| The type of each datum being sent or received in an Import or Export. More... | |
Public Member Functions | |
| virtual void | removeEmptyProcessesInPlace (const Teuchos::RCP< const map_type > &newMap) |
| Remove processes owning zero rows from the Map and their communicator. More... | |
| void | setCopyOrView (const Teuchos::DataAccess copyOrView) |
| Set whether this has copy (copyOrView = Teuchos::Copy) or view (copyOrView = Teuchos::View) semantics. More... | |
| Teuchos::DataAccess | getCopyOrView () const |
| Get whether this has copy (copyOrView = Teuchos::Copy) or view (copyOrView = Teuchos::View) semantics. More... | |
| void | assign (const MultiVector< Scalar, LocalOrdinal, GlobalOrdinal, Node > &src) |
Copy the contents of src into *this (deep copy). More... | |
| bool | isSameSize (const MultiVector< Scalar, LocalOrdinal, GlobalOrdinal, Node > &vec) const |
Constructors and destructor | |
| Vector () | |
| Default constructor: makes a Vector with no rows or columns. More... | |
| Vector (const Teuchos::RCP< const map_type > &map, const bool zeroOut=true) | |
| Basic constructor. More... | |
| Vector (const Vector< Scalar, LocalOrdinal, GlobalOrdinal, Node > &source, const Teuchos::DataAccess copyOrView) | |
| Copy constructor (shallow or deep copy). More... | |
| Vector (const Vector< Scalar, LocalOrdinal, GlobalOrdinal, Node > &source, const Teuchos::RCP< const map_type > &map, const local_ordinal_type rowOffset=0) | |
| "Offset view" constructor that views the input Vector's local data, but with the given Map, using the given row offset. More... | |
| Vector (const Teuchos::RCP< const map_type > &map, const Teuchos::ArrayView< const Scalar > &A) | |
| Set vector values from an existing array (copy) More... | |
| Vector (const Teuchos::RCP< const map_type > &map, const dual_view_type &view) | |
| Expert mode constructor, that takes a Kokkos::DualView of the Vector's data, and returns a Vector that views those data. More... | |
| Vector (const Teuchos::RCP< const map_type > &map, const dual_view_type &view, const dual_view_type &origView) | |
| Expert mode constructor, that takes a Kokkos::DualView of the Vector's data and the "original" Kokkos::DualView of the data, and returns a Vector that views those data. More... | |
| Vector (const MultiVector< Scalar, LocalOrdinal, GlobalOrdinal, Node > &X, const size_t j) | |
| Create a Vector that views a single column of the input MultiVector. More... | |
| Vector (const Vector< Scalar, LocalOrdinal, GlobalOrdinal, Node > &)=default | |
| Copy constructor (shallow copy). More... | |
| Vector (Vector< Scalar, LocalOrdinal, GlobalOrdinal, Node > &&)=default | |
| Move constructor (shallow move). More... | |
| Vector & | operator= (const Vector< Scalar, LocalOrdinal, GlobalOrdinal, Node > &)=default |
| Copy assignment (shallow copy). More... | |
| Vector< Scalar, LocalOrdinal, GlobalOrdinal, Node > & | operator= (Vector< Scalar, LocalOrdinal, GlobalOrdinal, Node > &&)=default |
| Move assigment (shallow move). More... | |
| virtual | ~Vector ()=default |
| Destructor (virtual for memory safety of derived classes). More... | |
Post-construction modification routines | |
| void | replaceGlobalValue (const GlobalOrdinal globalRow, const Scalar &value) const |
| Replace current value at the specified location with specified value. More... | |
| void | sumIntoGlobalValue (const GlobalOrdinal globalRow, const Scalar &value, const bool atomic=base_type::useAtomicUpdatesByDefault) const |
| Add value to existing value, using global (row) index. More... | |
| void | replaceLocalValue (const LocalOrdinal myRow, const Scalar &value) const |
| Replace current value at the specified location with specified values. More... | |
| void | sumIntoLocalValue (const LocalOrdinal myRow, const Scalar &value, const bool atomic=base_type::useAtomicUpdatesByDefault) const |
Add value to existing value, using local (row) index. More... | |
Extraction methods | |
| void | get1dCopy (const Teuchos::ArrayView< Scalar > &A) const |
| Return multi-vector values in user-provided two-dimensional array (using Teuchos memory management classes). More... | |
| Teuchos::ArrayRCP< Scalar > | getDataNonConst () |
| View of the local values of this vector. More... | |
| Teuchos::ArrayRCP< const Scalar > | getData () const |
| Const view of the local values of this vector. More... | |
|
Teuchos::RCP< const Vector < Scalar, LocalOrdinal, GlobalOrdinal, Node > > | offsetView (const Teuchos::RCP< const map_type > &subMap, const size_t offset) const |
|
Teuchos::RCP< Vector< Scalar, LocalOrdinal, GlobalOrdinal, Node > > | offsetViewNonConst (const Teuchos::RCP< const map_type > &subMap, const size_t offset) |
Mathematical methods | |
| dot_type | dot (const Vector< Scalar, LocalOrdinal, GlobalOrdinal, Node > &y) const |
| Return the dot product of this Vector and the input Vector x. More... | |
| mag_type | norm1 () const |
| Return the one-norm of this Vector. More... | |
| mag_type | norm2 () const |
| Return the two-norm of this Vector. More... | |
| mag_type | normInf () const |
| Return the infinity-norm of this Vector. More... | |
| Scalar | meanValue () const |
| Compute mean (average) value of this Vector. More... | |
Implementation of the Teuchos::Describable interface | |
| virtual std::string | description () const |
| Return a one-line description of this object. More... | |
| virtual void | describe (Teuchos::FancyOStream &out, const Teuchos::EVerbosityLevel verbLevel=Teuchos::Describable::verbLevel_default) const |
| Describe this object in a human-readable way to the given output stream. More... | |
Constructors and destructor | |
| void | swap (MultiVector< Scalar, LocalOrdinal, GlobalOrdinal, Node > &mv) |
| Return a deep copy of this MultiVector, with a different Node type. More... | |
Get a copy or view of a subset of rows and/or columns | |
The following methods get either a (deep) copy or a view (shallow copy) of a subset of rows and/or columns of the MultiVector. They return one of the following:
We prefer use of Kokkos classes to Teuchos Memory Management Classes. In particular, Teuchos::ArrayRCP reference counts are not thread safe, while Kokkos::View (and Kokkos::DualView) reference counts are thread safe. Not all of these methods are valid for a particular MultiVector. For instance, calling a method that accesses a view of the data in a 1-D format (i.e., get1dView) requires that the target MultiVector have constant stride. This category of methods also includes sync(), modify(), and getLocalView(), which help MultiVector implement DualView semantics. | |
| Teuchos::RCP< MultiVector < Scalar, LocalOrdinal, GlobalOrdinal, Node > > | subCopy (const Teuchos::Range1D &colRng) const |
| Return a MultiVector with copies of selected columns. More... | |
| Teuchos::RCP< MultiVector < Scalar, LocalOrdinal, GlobalOrdinal, Node > > | subCopy (const Teuchos::ArrayView< const size_t > &cols) const |
| Return a MultiVector with copies of selected columns. More... | |
| Teuchos::RCP< const MultiVector< Scalar, LocalOrdinal, GlobalOrdinal, Node > > | subView (const Teuchos::Range1D &colRng) const |
| Return a const MultiVector with const views of selected columns. More... | |
| Teuchos::RCP< const MultiVector< Scalar, LocalOrdinal, GlobalOrdinal, Node > > | subView (const Teuchos::ArrayView< const size_t > &cols) const |
| Return a const MultiVector with const views of selected columns. More... | |
| Teuchos::RCP< MultiVector < Scalar, LocalOrdinal, GlobalOrdinal, Node > > | subViewNonConst (const Teuchos::Range1D &colRng) |
| Return a MultiVector with views of selected columns. More... | |
| Teuchos::RCP< MultiVector < Scalar, LocalOrdinal, GlobalOrdinal, Node > > | subViewNonConst (const Teuchos::ArrayView< const size_t > &cols) |
| Return a MultiVector with views of selected columns. More... | |
| Teuchos::RCP< const Vector < Scalar, LocalOrdinal, GlobalOrdinal, Node > > | getVector (const size_t j) const |
| Return a Vector which is a const view of column j. More... | |
| Teuchos::RCP< Vector< Scalar, LocalOrdinal, GlobalOrdinal, Node > > | getVectorNonConst (const size_t j) |
| Return a Vector which is a nonconst view of column j. More... | |
| Teuchos::ArrayRCP< const Scalar > | getData (size_t j) const |
| Const view of the local values in a particular vector of this multivector. More... | |
| Teuchos::ArrayRCP< Scalar > | getDataNonConst (size_t j) |
| View of the local values in a particular vector of this multivector. More... | |
| void | get1dCopy (const Teuchos::ArrayView< Scalar > &A, const size_t LDA) const |
| Fill the given array with a copy of this multivector's local values. More... | |
| void | get2dCopy (const Teuchos::ArrayView< const Teuchos::ArrayView< Scalar > > &ArrayOfPtrs) const |
| Fill the given array with a copy of this multivector's local values. More... | |
| Teuchos::ArrayRCP< const Scalar > | get1dView () const |
| Const persisting (1-D) view of this multivector's local values. More... | |
| Teuchos::ArrayRCP < Teuchos::ArrayRCP< const Scalar > > | get2dView () const |
| Return const persisting pointers to values. More... | |
| Teuchos::ArrayRCP< Scalar > | get1dViewNonConst () |
| Nonconst persisting (1-D) view of this multivector's local values. More... | |
| Teuchos::ArrayRCP < Teuchos::ArrayRCP< Scalar > > | get2dViewNonConst () |
| Return non-const persisting pointers to values. More... | |
| void | clear_sync_state () |
| Clear "modified" flags on both host and device sides. More... | |
| template<class TargetDeviceType > | |
| void | sync () |
| Update data on device or host only if data in the other space has been marked as modified. More... | |
| void | sync_host () |
| Synchronize to Host. More... | |
| void | sync_device () |
| Synchronize to Device. More... | |
| template<class TargetDeviceType > | |
| bool | need_sync () const |
| Whether this MultiVector needs synchronization to the given space. More... | |
| bool | need_sync_host () const |
| Whether this MultiVector needs synchronization to the host. More... | |
| bool | need_sync_device () const |
| Whether this MultiVector needs synchronization to the device. More... | |
| template<class TargetDeviceType > | |
| void | modify () |
Mark data as modified on the given device TargetDeviceType. More... | |
| void | modify_device () |
| Mark data as modified on the device side. More... | |
| void | modify_host () |
| Mark data as modified on the host side. More... | |
| template<class TargetDeviceType > | |
| Kokkos::Impl::if_c < std::is_same< typename device_type::memory_space, typename TargetDeviceType::memory_space > ::value, typename dual_view_type::t_dev, typename dual_view_type::t_host >::type | getLocalView () const |
| Return a view of the local data on a specific device. More... | |
| dual_view_type::t_host | getLocalViewHost () const |
| A local Kokkos::View of host memory. More... | |
| dual_view_type::t_dev | getLocalViewDevice () const |
| A local Kokkos::View of device memory. More... | |
Mathematical methods | |
| void | dot (const MultiVector< Scalar, LocalOrdinal, GlobalOrdinal, Node > &A, const Teuchos::ArrayView< dot_type > &dots) const |
| Compute the dot product of each corresponding pair of vectors (columns) in A and B. More... | |
| template<typename T > | |
| std::enable_if< !(std::is_same < dot_type, T >::value), void > ::type | dot (const MultiVector< Scalar, LocalOrdinal, GlobalOrdinal, Node > &A, const Teuchos::ArrayView< T > &dots) const |
| Compute the dot product of each corresponding pair of vectors (columns) in A and B. More... | |
| template<typename T > | |
| std::enable_if< !(std::is_same < dot_type, T >::value), void > ::type | dot (const MultiVector< Scalar, LocalOrdinal, GlobalOrdinal, Node > &A, std::vector< T > &dots) const |
| Like the above dot() overload, but for std::vector output. More... | |
| void | dot (const MultiVector< Scalar, LocalOrdinal, GlobalOrdinal, Node > &A, const Kokkos::View< dot_type *, Kokkos::HostSpace > &norms) const |
| Compute the dot product of each corresponding pair of vectors (columns) in A and B, storing the result in a device View. More... | |
| template<class ViewType > | |
| void | dot (typename std::enable_if< std::is_same< typename ViewType::value_type, dot_type >::value &&std::is_same< typename ViewType::memory_space, typename device_type::memory_space >::value, const MultiVector< Scalar, LocalOrdinal, GlobalOrdinal, Node >>::type &A, const ViewType &dots) const |
| template<typename T > | |
| std::enable_if< !(std::is_same < dot_type, T >::value), void > ::type | dot (const MultiVector< Scalar, LocalOrdinal, GlobalOrdinal, Node > &A, const Kokkos::View< T *, device_type > &dots) const |
| Compute the dot product of each corresponding pair of vectors (columns) in A and B, storing the result in a device view. More... | |
| void | abs (const MultiVector< Scalar, LocalOrdinal, GlobalOrdinal, Node > &A) |
| Put element-wise absolute values of input Multi-vector in target: A = abs(this) More... | |
| void | reciprocal (const MultiVector< Scalar, LocalOrdinal, GlobalOrdinal, Node > &A) |
| Put element-wise reciprocal values of input Multi-vector in target, this(i,j) = 1/A(i,j). More... | |
| void | scale (const Scalar &alpha) |
Scale in place: this = alpha*this. More... | |
| void | scale (const Teuchos::ArrayView< const Scalar > &alpha) |
Scale each column in place: this[j] = alpha[j]*this[j]. More... | |
| void | scale (const Kokkos::View< const impl_scalar_type *, device_type > &alpha) |
Scale each column in place: this[j] = alpha[j]*this[j]. More... | |
| void | scale (const Scalar &alpha, const MultiVector< Scalar, LocalOrdinal, GlobalOrdinal, Node > &A) |
Scale in place: this = alpha * A. More... | |
| void | update (const Scalar &alpha, const MultiVector< Scalar, LocalOrdinal, GlobalOrdinal, Node > &A, const Scalar &beta) |
Update: this = beta*this + alpha*A. More... | |
| void | update (const Scalar &alpha, const MultiVector< Scalar, LocalOrdinal, GlobalOrdinal, Node > &A, const Scalar &beta, const MultiVector< Scalar, LocalOrdinal, GlobalOrdinal, Node > &B, const Scalar &gamma) |
Update: this = gamma*this + alpha*A + beta*B. More... | |
| void | norm1 (const Kokkos::View< mag_type *, Kokkos::HostSpace > &norms) const |
| Compute the one-norm of each vector (column), storing the result in a host view. More... | |
| template<class ViewType > | |
| std::enable_if< std::is_same < typename ViewType::value_type, mag_type > ::value &&std::is_same < typename ViewType::memory_space, typename device_type::memory_space > ::value >::type | norm1 (const ViewType &norms) const |
| template<typename T > | |
| std::enable_if< !(std::is_same < mag_type, T >::value), void > ::type | norm1 (const Kokkos::View< T *, device_type > &norms) const |
| Compute the one-norm of each vector (column), storing the result in a device view. More... | |
| void | norm1 (const Teuchos::ArrayView< mag_type > &norms) const |
| Compute the one-norm of each vector (column). More... | |
| template<typename T > | |
| std::enable_if< !(std::is_same < mag_type, T >::value), void > ::type | norm1 (const Teuchos::ArrayView< T > &norms) const |
| Compute the one-norm of each vector (column). More... | |
| void | norm2 (const Kokkos::View< mag_type *, Kokkos::HostSpace > &norms) const |
| Compute the two-norm of each vector (column), storing the result in a host View. More... | |
| template<class ViewType > | |
| std::enable_if< std::is_same < typename ViewType::value_type, mag_type > ::value &&std::is_same < typename ViewType::memory_space, typename device_type::memory_space > ::value >::type | norm2 (const ViewType &norms) const |
| template<typename T > | |
| std::enable_if< !(std::is_same < mag_type, T >::value), void > ::type | norm2 (const Kokkos::View< T *, device_type > &norms) const |
| Compute the two-norm of each vector (column), storing the result in a device view. More... | |
| void | norm2 (const Teuchos::ArrayView< mag_type > &norms) const |
| Compute the two-norm of each vector (column). More... | |
| template<typename T > | |
| std::enable_if< !(std::is_same < mag_type, T >::value), void > ::type | norm2 (const Teuchos::ArrayView< T > &norms) const |
| Compute the two-norm of each vector (column). More... | |
| void | normInf (const Kokkos::View< mag_type *, Kokkos::HostSpace > &norms) const |
| Compute the infinity-norm of each vector (column), storing the result in a host View. More... | |
| template<class ViewType > | |
| std::enable_if< std::is_same < typename ViewType::value_type, mag_type > ::value &&std::is_same < typename ViewType::memory_space, typename device_type::memory_space > ::value >::type | normInf (const ViewType &norms) const |
| template<typename T > | |
| std::enable_if< !(std::is_same < mag_type, T >::value), void > ::type | normInf (const Kokkos::View< T *, device_type > &norms) const |
| Compute the infinity-norm of each vector (column), storing the result in a device view. More... | |
| void | normInf (const Teuchos::ArrayView< mag_type > &norms) const |
| Compute the infinity-norm of each vector (column), storing the result in a Teuchos::ArrayView. More... | |
| template<typename T > | |
| std::enable_if< !(std::is_same < mag_type, T >::value), void > ::type | normInf (const Teuchos::ArrayView< T > &norms) const |
| Compute the infinity-norm of each vector (column), storing the result in a Teuchos::ArrayView. More... | |
| void | meanValue (const Teuchos::ArrayView< impl_scalar_type > &means) const |
| Compute mean (average) value of each column. More... | |
| template<typename T > | |
| std::enable_if<!std::is_same < impl_scalar_type, T >::value, void >::type | meanValue (const Teuchos::ArrayView< T > &means) const |
| void | multiply (Teuchos::ETransp transA, Teuchos::ETransp transB, const Scalar &alpha, const MultiVector< Scalar, LocalOrdinal, GlobalOrdinal, Node > &A, const MultiVector< Scalar, LocalOrdinal, GlobalOrdinal, Node > &B, const Scalar &beta) |
Matrix-matrix multiplication: this = beta*this + alpha*op(A)*op(B). More... | |
| void | elementWiseMultiply (Scalar scalarAB, const Vector< Scalar, LocalOrdinal, GlobalOrdinal, Node > &A, const MultiVector< Scalar, LocalOrdinal, GlobalOrdinal, Node > &B, Scalar scalarThis) |
| Multiply a Vector A elementwise by a MultiVector B. More... | |
Attribute access functions | |
| size_t | getNumVectors () const |
| Number of columns in the multivector. More... | |
| size_t | getLocalLength () const |
| Local number of rows on the calling process. More... | |
| global_size_t | getGlobalLength () const |
| Global number of rows in the multivector. More... | |
| size_t | getStride () const |
| Stride between columns in the multivector. More... | |
| bool | isConstantStride () const |
| Whether this multivector has constant stride between columns. More... | |
Methods for use only by experts | |
| virtual void | removeEmptyProcessesInPlace (const Teuchos::RCP< const map_type > &newMap) |
| Remove processes which contain no entries in this object's Map. More... | |
Public methods for redistributing data | |
| void | doImport (const SrcDistObject &source, const Import< LocalOrdinal, GlobalOrdinal, Node > &importer, const CombineMode CM, const bool restrictedMode=false) |
| Import data into this object using an Import object ("forward mode"). More... | |
| void | doImport (const SrcDistObject &source, const Export< LocalOrdinal, GlobalOrdinal, Node > &exporter, const CombineMode CM, const bool restrictedMode=false) |
| Import data into this object using an Export object ("reverse mode"). More... | |
| void | doExport (const SrcDistObject &source, const Export< LocalOrdinal, GlobalOrdinal, Node > &exporter, const CombineMode CM, const bool restrictedMode=false) |
| Export data into this object using an Export object ("forward mode"). More... | |
| void | doExport (const SrcDistObject &source, const Import< LocalOrdinal, GlobalOrdinal, Node > &importer, const CombineMode CM, const bool restrictedMode=false) |
| Export data into this object using an Import object ("reverse mode"). More... | |
Attribute accessor methods | |
| bool | isDistributed () const |
| Whether this is a globally distributed object. More... | |
| virtual Teuchos::RCP< const map_type > | getMap () const |
| The Map describing the parallel distribution of this object. More... | |
I/O methods | |
| void | print (std::ostream &os) const |
| Print this object to the given output stream. More... | |
Protected Member Functions | |
| virtual void | doTransfer (const SrcDistObject &src, const ::Tpetra::Details::Transfer< local_ordinal_type, global_ordinal_type, node_type > &transfer, const char modeString[], const ReverseOption revOp, const CombineMode CM, const bool restrictedMode) |
| Redistribute data across (MPI) processes. More... | |
| virtual bool | reallocArraysForNumPacketsPerLid (const size_t numExportLIDs, const size_t numImportLIDs) |
| Reallocate numExportPacketsPerLID_ and/or numImportPacketsPerLID_, if necessary. More... | |
| virtual void | doTransferNew (const SrcDistObject &src, const CombineMode CM, const size_t numSameIDs, const Kokkos::DualView< const local_ordinal_type *, buffer_device_type > &permuteToLIDs, const Kokkos::DualView< const local_ordinal_type *, buffer_device_type > &permuteFromLIDs, const Kokkos::DualView< const local_ordinal_type *, buffer_device_type > &remoteLIDs, const Kokkos::DualView< const local_ordinal_type *, buffer_device_type > &exportLIDs, Distributor &distor, const ReverseOption revOp, const bool commOnHost, const bool restrictedMode) |
| Implementation detail of doTransfer. More... | |
Misc. implementation details | |
| std::string | descriptionImpl (const std::string &className) const |
| Implementation of description() for this class, and its subclass Vector. More... | |
| std::string | localDescribeToString (const Teuchos::EVerbosityLevel vl) const |
| Print the calling process' verbose describe() information to the returned string. More... | |
| void | describeImpl (Teuchos::FancyOStream &out, const std::string &className, const Teuchos::EVerbosityLevel verbLevel=Teuchos::Describable::verbLevel_default) const |
| Implementation of describe() for this class, and its subclass Vector. More... | |
| bool | vectorIndexOutOfRange (const size_t VectorIndex) const |
| template<class T > | |
| Teuchos::ArrayRCP< T > | getSubArrayRCP (Teuchos::ArrayRCP< T > arr, size_t j) const |
| Persisting view of j-th column in the given ArrayRCP. More... | |
| size_t | getOrigNumLocalRows () const |
| "Original" number of rows in the (local) data. More... | |
| size_t | getOrigNumLocalCols () const |
| "Original" number of columns in the (local) data. More... | |
Protected Attributes | |
| dual_view_type | view_ |
| The Kokkos::DualView containing the MultiVector's data. More... | |
| dual_view_type | origView_ |
| The "original view" of the MultiVector's data. More... | |
| Teuchos::Array< size_t > | whichVectors_ |
| Indices of columns this multivector is viewing. More... | |
Related Functions | |
(Note that these are not member functions.) | |
| template<class Scalar , class LocalOrdinal , class GlobalOrdinal , class Node > | |
| Teuchos::RCP< MultiVector < Scalar, LocalOrdinal, GlobalOrdinal, Node > > | createMultiVector (const Teuchos::RCP< const Map< LocalOrdinal, GlobalOrdinal, Node > > &map, const size_t numVectors) |
| Nonmember MultiVector "constructor": Create a MultiVector from a given Map. More... | |
| template<class Scalar , class LocalOrdinal , class GlobalOrdinal , class Node > | |
| Vector< Scalar, LocalOrdinal, GlobalOrdinal, Node > | createCopy (const Vector< Scalar, LocalOrdinal, GlobalOrdinal, Node > &src) |
| Return a deep copy of the given Vector. More... | |
| template<class Scalar , class LocalOrdinal , class GlobalOrdinal , class Node > | |
| Teuchos::RCP< Vector< Scalar, LocalOrdinal, GlobalOrdinal, Node > > | createVector (const Teuchos::RCP< const Map< LocalOrdinal, GlobalOrdinal, Node > > &map) |
| Nonmember Vector "constructor": Create a Vector from a given Map. More... | |
| template<class DS , class DL , class DG , class DN , class SS , class SL , class SG , class SN > | |
| void | deep_copy (MultiVector< DS, DL, DG, DN > &dst, const MultiVector< SS, SL, SG, SN > &src) |
Copy the contents of the MultiVector src into dst. More... | |
| template<class ST , class LO , class GO , class NT > | |
| MultiVector< ST, LO, GO, NT > | createCopy (const MultiVector< ST, LO, GO, NT > &src) |
| Return a deep copy of the given MultiVector. More... | |
Methods implemented by subclasses and used by doTransfer(). | |
The doTransfer() method uses the subclass' implementations of these methods to implement data transfer. Subclasses of DistObject must implement these methods. This is an instance of the Template Method Pattern. ("Template" here doesn't mean "C++ template"; it means "pattern with holes that are filled in by the subclass' method implementations.") | |
| Teuchos::RCP< const map_type > | map_ |
| The Map over which this object is distributed. More... | |
| Kokkos::DualView< packet_type *, buffer_device_type > | imports_ |
| Buffer into which packed data are imported (received from other processes). More... | |
| Kokkos::DualView< size_t *, buffer_device_type > | numImportPacketsPerLID_ |
| Number of packets to receive for each receive operation. More... | |
| Kokkos::DualView< packet_type *, buffer_device_type > | exports_ |
| Buffer from which packed data are exported (sent to other processes). More... | |
| Kokkos::DualView< size_t *, buffer_device_type > | numExportPacketsPerLID_ |
| Number of packets to send for each send operation. More... | |
| virtual void | copyAndPermute (const SrcDistObject &source, const size_t numSameIDs, const Kokkos::DualView< const local_ordinal_type *, buffer_device_type > &permuteToLIDs, const Kokkos::DualView< const local_ordinal_type *, buffer_device_type > &permuteFromLIDs) |
| Perform copies and permutations that are local to the calling (MPI) process. More... | |
| virtual void | packAndPrepare (const SrcDistObject &source, const Kokkos::DualView< const local_ordinal_type *, buffer_device_type > &exportLIDs, Kokkos::DualView< packet_type *, buffer_device_type > &exports, Kokkos::DualView< size_t *, buffer_device_type > numPacketsPerLID, size_t &constantNumPackets, Distributor &distor) |
| Pack data and metadata for communication (sends). More... | |
| virtual void | unpackAndCombine (const Kokkos::DualView< const local_ordinal_type *, buffer_device_type > &importLIDs, Kokkos::DualView< packet_type *, buffer_device_type > imports, Kokkos::DualView< size_t *, buffer_device_type > numPacketsPerLID, const size_t constantNumPackets, Distributor &distor, const CombineMode combineMode) |
| Perform any unpacking and combining after communication. More... | |
| bool | reallocImportsIfNeeded (const size_t newSize, const bool verbose, const std::string *prefix) |
| Reallocate imports_ if needed. More... | |
Post-construction modification routines | |
| void | replaceGlobalValue (const GlobalOrdinal gblRow, const size_t col, const impl_scalar_type &value) const |
| Replace value in host memory, using global row index. More... | |
| template<typename T > | |
| std::enable_if<!std::is_same < T, impl_scalar_type >::value &&std::is_convertible< T, impl_scalar_type >::value, void >::type | replaceGlobalValue (GlobalOrdinal globalRow, size_t col, const T &value) const |
| Like the above replaceGlobalValue, but only enabled if T differs from impl_scalar_type. More... | |
| void | sumIntoGlobalValue (const GlobalOrdinal gblRow, const size_t col, const impl_scalar_type &value, const bool atomic=useAtomicUpdatesByDefault) const |
| Update (+=) a value in host memory, using global row index. More... | |
| template<typename T > | |
| std::enable_if<!std::is_same < T, impl_scalar_type >::value &&std::is_convertible< T, impl_scalar_type >::value, void >::type | sumIntoGlobalValue (const GlobalOrdinal gblRow, const size_t col, const T &val, const bool atomic=useAtomicUpdatesByDefault) const |
| Like the above sumIntoGlobalValue, but only enabled if T differs from impl_scalar_type. More... | |
| void | replaceLocalValue (const LocalOrdinal lclRow, const size_t col, const impl_scalar_type &value) const |
| Replace value in host memory, using local (row) index. More... | |
| template<typename T > | |
| std::enable_if<!std::is_same < T, impl_scalar_type >::value &&std::is_convertible< T, impl_scalar_type >::value, void >::type | replaceLocalValue (const LocalOrdinal lclRow, const size_t col, const T &val) const |
| Like the above replaceLocalValue, but only enabled if T differs from impl_scalar_type. More... | |
| void | sumIntoLocalValue (const LocalOrdinal lclRow, const size_t col, const impl_scalar_type &val, const bool atomic=useAtomicUpdatesByDefault) const |
| Update (+=) a value in host memory, using local row index. More... | |
| template<typename T > | |
| std::enable_if<!std::is_same < T, impl_scalar_type >::value &&std::is_convertible< T, impl_scalar_type >::value, void >::type | sumIntoLocalValue (const LocalOrdinal lclRow, const size_t col, const T &val, const bool atomic=useAtomicUpdatesByDefault) const |
| Like the above sumIntoLocalValue, but only enabled if T differs from impl_scalar_type. More... | |
| void | putScalar (const Scalar &value) |
| Set all values in the multivector with the given value. More... | |
| template<typename T > | |
| std::enable_if<!std::is_same < T, impl_scalar_type >::value &&std::is_convertible< T, impl_scalar_type >::value, void >::type | putScalar (const T &value) |
| Set all values in the multivector with the given value. More... | |
| void | randomize () |
| Set all values in the multivector to pseudorandom numbers. More... | |
| void | randomize (const Scalar &minVal, const Scalar &maxVal) |
| Set all values in the multivector to pseudorandom numbers in the given range. More... | |
| void | replaceMap (const Teuchos::RCP< const map_type > &map) |
| Replace the underlying Map in place. More... | |
| void | reduce () |
| Sum values of a locally replicated multivector across all processes. More... | |
| static const bool | useAtomicUpdatesByDefault |
| Whether sumIntoLocalValue and sumIntoGlobalValue should use atomic updates by default. More... | |
Implementation of Tpetra::DistObject | |
| using | buffer_device_type = typename DistObject< scalar_type, local_ordinal_type, global_ordinal_type, node_type >::buffer_device_type |
| Kokkos::Device specialization for communication buffers. More... | |
| virtual bool | checkSizes (const SrcDistObject &sourceObj) |
Whether data redistribution between sourceObj and this object is legal. More... | |
| virtual size_t | constantNumberOfPackets () const |
| Number of packets to send per LID. More... | |
| virtual void | copyAndPermute (const SrcDistObject &sourceObj, const size_t numSameIDs, const Kokkos::DualView< const local_ordinal_type *, buffer_device_type > &permuteToLIDs, const Kokkos::DualView< const local_ordinal_type *, buffer_device_type > &permuteFromLIDs) |
| virtual void | packAndPrepare (const SrcDistObject &sourceObj, const Kokkos::DualView< const local_ordinal_type *, buffer_device_type > &exportLIDs, Kokkos::DualView< impl_scalar_type *, buffer_device_type > &exports, Kokkos::DualView< size_t *, buffer_device_type >, size_t &constantNumPackets, Distributor &) |
| virtual void | unpackAndCombine (const Kokkos::DualView< const local_ordinal_type *, buffer_device_type > &importLIDs, Kokkos::DualView< impl_scalar_type *, buffer_device_type > imports, Kokkos::DualView< size_t *, buffer_device_type >, const size_t constantNumPackets, Distributor &, const CombineMode CM) |
A distributed dense vector.
| Scalar | The type of each entry of the vector. (You can use real-valued or complex-valued types here, unlike in Epetra, where the scalar type is always double.) |
| LocalOrdinal | The type of local indices. See the documentation of Map for requirements. |
| GlobalOrdinal | The type of global indices. See the documentation of Map for requirements. |
| Node | The Kokkos Node type. See the documentation of Map for requirements. |
This class inherits from MultiVector, and has the same template parameters. A Vector is a special case of a MultiVector that has only one vector (column). It may be used wherever a MultiVector may be used. Please see the documentation of MultiVector for more details.
Definition at line 81 of file Tpetra_Vector_decl.hpp.
| typedef Scalar Tpetra::Vector< Scalar, LocalOrdinal, GlobalOrdinal, Node >::scalar_type |
This class' first template parameter; the type of each entry in the Vector.
Definition at line 92 of file Tpetra_Vector_decl.hpp.
| typedef base_type::impl_scalar_type Tpetra::Vector< Scalar, LocalOrdinal, GlobalOrdinal, Node >::impl_scalar_type |
The type used internally in place of Scalar.
Some Scalar types might not work with Kokkos on all execution spaces, due to missing CUDA device macros or volatile overloads. The C++ standard type std::complex<T> has this problem. To fix this, we replace std::complex<T> values internally with the (usually) bitwise identical type Kokkos::complex<T>. The latter is the impl_scalar_type corresponding to Scalar = std::complex.
Definition at line 102 of file Tpetra_Vector_decl.hpp.
| typedef LocalOrdinal Tpetra::Vector< Scalar, LocalOrdinal, GlobalOrdinal, Node >::local_ordinal_type |
This class' second template parameter; the type of local indices.
Definition at line 104 of file Tpetra_Vector_decl.hpp.
| typedef GlobalOrdinal Tpetra::Vector< Scalar, LocalOrdinal, GlobalOrdinal, Node >::global_ordinal_type |
This class' third template parameter; the type of global indices.
Definition at line 106 of file Tpetra_Vector_decl.hpp.
| typedef Node::device_type Tpetra::Vector< Scalar, LocalOrdinal, GlobalOrdinal, Node >::device_type |
The Kokkos device type.
Definition at line 108 of file Tpetra_Vector_decl.hpp.
| typedef Node Tpetra::Vector< Scalar, LocalOrdinal, GlobalOrdinal, Node >::node_type |
The Kokkos Node type.
Definition at line 111 of file Tpetra_Vector_decl.hpp.
| typedef base_type::dot_type Tpetra::Vector< Scalar, LocalOrdinal, GlobalOrdinal, Node >::dot_type |
Type of an inner ("dot") product result.
This is usually the same as impl_scalar_type, but may differ if impl_scalar_type is e.g., an uncertainty quantification type from the Stokhos package.
Definition at line 118 of file Tpetra_Vector_decl.hpp.
| typedef base_type::mag_type Tpetra::Vector< Scalar, LocalOrdinal, GlobalOrdinal, Node >::mag_type |
Type of a norm result.
This is usually the same as the type of the magnitude (absolute value) of impl_scalar_type, but may differ if impl_scalar_type is e.g., an uncertainty quantification type from the Stokhos package.
Definition at line 126 of file Tpetra_Vector_decl.hpp.
| typedef base_type::dual_view_type Tpetra::Vector< Scalar, LocalOrdinal, GlobalOrdinal, Node >::dual_view_type |
Kokkos::DualView specialization used by this class.
Definition at line 129 of file Tpetra_Vector_decl.hpp.
| typedef base_type::map_type Tpetra::Vector< Scalar, LocalOrdinal, GlobalOrdinal, Node >::map_type |
The type of the Map specialization used by this class.
Definition at line 132 of file Tpetra_Vector_decl.hpp.
|
inherited |
Type of the (new) Kokkos execution space.
The execution space implements parallel operations, like parallel_for, parallel_reduce, and parallel_scan.
Definition at line 481 of file Tpetra_MultiVector_decl.hpp.
|
protectedinherited |
Kokkos::Device specialization for communication buffers.
See #1088 for why this is not just device_type::device_type.
Definition at line 2435 of file Tpetra_MultiVector_decl.hpp.
|
inherited |
The type of each datum being sent or received in an Import or Export.
Note that this type does not always correspond to the Scalar template parameter of subclasses.
Definition at line 338 of file Tpetra_DistObject_decl.hpp.
| Tpetra::Vector< Scalar, LocalOrdinal, GlobalOrdinal, Node >::Vector | ( | ) |
Default constructor: makes a Vector with no rows or columns.
Definition at line 63 of file Tpetra_Vector_def.hpp.
|
explicit |
Basic constructor.
| map | [in] The Vector's Map. The Map describes the distribution of rows over process(es) in the Map's communicator. |
| zeroOut | [in] If true (the default), require that all the Vector's entries be zero on return. If false, the Vector's entries have undefined values on return, and must be set explicitly. |
Definition at line 69 of file Tpetra_Vector_def.hpp.
| Tpetra::Vector< Scalar, LocalOrdinal, GlobalOrdinal, Node >::Vector | ( | const Vector< Scalar, LocalOrdinal, GlobalOrdinal, Node > & | source, |
| const Teuchos::DataAccess | copyOrView | ||
| ) |
Copy constructor (shallow or deep copy).
| source | [in] The Vector to copy. |
| copyOrView | [in] If Teuchos::View, return a shallow copy (a view) of source. If Teuchos::Copy, return a deep copy of source. Regardless, the result has "view semantics." This means that copy construction or assignment (operator=) with the resulting object will always do a shallow copy, and will transmit view semantics to the result of the shallow copy. |
Definition at line 76 of file Tpetra_Vector_def.hpp.
| Tpetra::Vector< Scalar, LocalOrdinal, GlobalOrdinal, Node >::Vector | ( | const Vector< Scalar, LocalOrdinal, GlobalOrdinal, Node > & | source, |
| const Teuchos::RCP< const map_type > & | map, | ||
| const local_ordinal_type | rowOffset = 0 |
||
| ) |
| Tpetra::Vector< Scalar, LocalOrdinal, GlobalOrdinal, Node >::Vector | ( | const Teuchos::RCP< const map_type > & | map, |
| const Teuchos::ArrayView< const Scalar > & | A | ||
| ) |
Set vector values from an existing array (copy)
Definition at line 91 of file Tpetra_Vector_def.hpp.
| Tpetra::Vector< Scalar, LocalOrdinal, GlobalOrdinal, Node >::Vector | ( | const Teuchos::RCP< const map_type > & | map, |
| const dual_view_type & | view | ||
| ) |
Expert mode constructor, that takes a Kokkos::DualView of the Vector's data, and returns a Vector that views those data.
See the documentation of the MultiVector (parent class) constructor that takes the same arguments.
| map | [in] Map describing the distribution of rows. |
| view | [in] View of the data (shallow copy). |
Definition at line 98 of file Tpetra_Vector_def.hpp.
| Tpetra::Vector< Scalar, LocalOrdinal, GlobalOrdinal, Node >::Vector | ( | const Teuchos::RCP< const map_type > & | map, |
| const dual_view_type & | view, | ||
| const dual_view_type & | origView | ||
| ) |
Expert mode constructor, that takes a Kokkos::DualView of the Vector's data and the "original" Kokkos::DualView of the data, and returns a Vector that views those data.
See the documentation of the MultiVector (parent class) constructor that takes the same arguments.
| map | [in] Map describing the distribution of rows. |
| view | [in] View of the data (shallow copy). |
| origView | [in] "Original" view of the data (shallow copy). |
Definition at line 105 of file Tpetra_Vector_def.hpp.
| Tpetra::Vector< Scalar, LocalOrdinal, GlobalOrdinal, Node >::Vector | ( | const MultiVector< Scalar, LocalOrdinal, GlobalOrdinal, Node > & | X, |
| const size_t | j | ||
| ) |
Create a Vector that views a single column of the input MultiVector.
| X | [in] Input MultiVector to view (in possibly nonconst fashion). |
| j | [in] The column of X to view. |
Definition at line 113 of file Tpetra_Vector_def.hpp.
|
default |
Copy constructor (shallow copy).
Vector's copy constructor always does a shallow copy. Use the nonmember function Tpetra::deep_copy (see Tpetra_MultiVector_decl.hpp) to deep-copy one existing Vector to another, and use the two-argument "copy constructor" below (with copyOrView=Teuchos::Copy) to create a Vector that is a deep copy of an existing Vector.
|
default |
Move constructor (shallow move).
|
virtualdefault |
Destructor (virtual for memory safety of derived classes).
=default declarations for copy construction, move construction, copy assignment, and move assignment.
|
default |
Copy assignment (shallow copy).
Vector's copy assignment operator always does a shallow copy. Use the nonmember function Tpetra::deep_copy (see Tpetra_MultiVector_decl.hpp) to deep-copy one existing Vector to another.
|
default |
Move assigment (shallow move).
| void Tpetra::Vector< Scalar, LocalOrdinal, GlobalOrdinal, Node >::replaceGlobalValue | ( | const GlobalOrdinal | globalRow, |
| const Scalar & | value | ||
| ) | const |
Replace current value at the specified location with specified value.
globalRow must be a valid global element on this node, according to the row map. Definition at line 121 of file Tpetra_Vector_def.hpp.
| void Tpetra::Vector< Scalar, LocalOrdinal, GlobalOrdinal, Node >::sumIntoGlobalValue | ( | const GlobalOrdinal | globalRow, |
| const Scalar & | value, | ||
| const bool | atomic = base_type::useAtomicUpdatesByDefault |
||
| ) | const |
Add value to existing value, using global (row) index.
Add the given value to the existing value at row globalRow (a global index).
This method affects the host memory version of the data. If the DeviceType template parameter is a device that has two memory spaces, and you want to modify the non-host version of the data, you must access the device View directly by calling getLocalView(). Please see modify(), sync(), and the discussion of DualView semantics elsewhere in the documentation.
| globalRow | [in] Global row index of the entry to modify. This must be a valid global row index on the calling process with respect to the Vector's Map. |
| value | [in] Incoming value to add to the entry. |
| atomic | [in] Whether to use an atomic update. If this class' execution space is not Kokkos::Serial, then this is true by default, else it is false by default. |
Definition at line 128 of file Tpetra_Vector_def.hpp.
| void Tpetra::Vector< Scalar, LocalOrdinal, GlobalOrdinal, Node >::replaceLocalValue | ( | const LocalOrdinal | myRow, |
| const Scalar & | value | ||
| ) | const |
Replace current value at the specified location with specified values.
localRow must be a valid local element on this node, according to the row map. Definition at line 138 of file Tpetra_Vector_def.hpp.
| void Tpetra::Vector< Scalar, LocalOrdinal, GlobalOrdinal, Node >::sumIntoLocalValue | ( | const LocalOrdinal | myRow, |
| const Scalar & | value, | ||
| const bool | atomic = base_type::useAtomicUpdatesByDefault |
||
| ) | const |
Add value to existing value, using local (row) index.
Add the given value to the existing value at row localRow (a local index).
This method affects the host memory version of the data. If the DeviceType template parameter is a device that has two memory spaces, and you want to modify the non-host version of the data, you must access the device View directly by calling getLocalView(). Please see modify(), sync(), and the discussion of DualView semantics elsewhere in the documentation.
| localRow | [in] Local row index of the entry to modify. |
| value | [in] Incoming value to add to the entry. |
| atomic | [in] Whether to use an atomic update. If this class' execution space is not Kokkos::Serial, then this is true by default, else it is false by default. |
Definition at line 145 of file Tpetra_Vector_def.hpp.
| void Tpetra::Vector< Scalar, LocalOrdinal, GlobalOrdinal, Node >::get1dCopy | ( | const Teuchos::ArrayView< Scalar > & | A | ) | const |
Return multi-vector values in user-provided two-dimensional array (using Teuchos memory management classes).
Definition at line 155 of file Tpetra_Vector_def.hpp.
|
inline |
View of the local values of this vector.
Definition at line 344 of file Tpetra_Vector_decl.hpp.
|
inline |
Const view of the local values of this vector.
Definition at line 350 of file Tpetra_Vector_decl.hpp.
| Vector< Scalar, LocalOrdinal, GlobalOrdinal, Node >::dot_type Tpetra::Vector< Scalar, LocalOrdinal, GlobalOrdinal, Node >::dot | ( | const Vector< Scalar, LocalOrdinal, GlobalOrdinal, Node > & | y | ) | const |
Return the dot product of this Vector and the input Vector x.
Definition at line 163 of file Tpetra_Vector_def.hpp.
| Vector< Scalar, LocalOrdinal, GlobalOrdinal, Node >::mag_type Tpetra::Vector< Scalar, LocalOrdinal, GlobalOrdinal, Node >::norm1 | ( | ) | const |
Return the one-norm of this Vector.
Definition at line 183 of file Tpetra_Vector_def.hpp.
| Vector< Scalar, LocalOrdinal, GlobalOrdinal, Node >::mag_type Tpetra::Vector< Scalar, LocalOrdinal, GlobalOrdinal, Node >::norm2 | ( | ) | const |
Return the two-norm of this Vector.
Definition at line 193 of file Tpetra_Vector_def.hpp.
| Vector< Scalar, LocalOrdinal, GlobalOrdinal, Node >::mag_type Tpetra::Vector< Scalar, LocalOrdinal, GlobalOrdinal, Node >::normInf | ( | ) | const |
Return the infinity-norm of this Vector.
Definition at line 203 of file Tpetra_Vector_def.hpp.
| Scalar Tpetra::Vector< Scalar, LocalOrdinal, GlobalOrdinal, Node >::meanValue | ( | ) | const |
Compute mean (average) value of this Vector.
Definition at line 173 of file Tpetra_Vector_def.hpp.
|
virtual |
Return a one-line description of this object.
Reimplemented from Tpetra::MultiVector< Scalar, LocalOrdinal, GlobalOrdinal, Node >.
Definition at line 272 of file Tpetra_Vector_def.hpp.
|
virtual |
Describe this object in a human-readable way to the given output stream.
You must call this method as a collective over all processes in this object's communicator.
| out | [out] Output stream to which to write. Only Process 0 in this object's communicator may write to the output stream. |
| verbLevel | [in] Verbosity level. This also controls whether this method does any communication. At verbosity levels higher (greater) than Teuchos::VERB_LOW, this method may behave as a collective over the object's communicator. |
Teuchos::FancyOStream wraps std::ostream. It adds features like tab levels. If you just want to wrap std::cout, try this:
Reimplemented from Tpetra::MultiVector< Scalar, LocalOrdinal, GlobalOrdinal, Node >.
Definition at line 279 of file Tpetra_Vector_def.hpp.
|
inherited |
Return a deep copy of this MultiVector, with a different Node type.
| node2 | [in/out] The new Node type. |
mv with contents of *this. Definition at line 4555 of file Tpetra_MultiVector_def.hpp.
|
inherited |
Replace value in host memory, using global row index.
Replace the current value at row gblRow (a global index) and column col with the given value. The column index is zero based.
This method affects the host memory version of the data. If device_type is a Kokkos device that has two memory spaces, and you want to modify the non-host version of the data, you must access the device View directly by calling getLocalView(). Please see modify(), sync(), and the discussion of DualView semantics elsewhere in the documentation. You are responsible for calling modify() and sync(), if needed; this method doesn't do that.
This method does not have an "atomic" option like sumIntoGlobalValue. This is deliberate. Replacement is not commutative, unlike += (modulo rounding error). Concurrent calls to replaceGlobalValue on different threads that modify the same entry/ies have undefined results. (It's not just that one thread might win; it's that the value might get messed up.)
| gblRow | [in] Global row index of the entry to modify. This must be a valid global row index on the calling process with respect to the MultiVector's Map. |
| col | [in] Column index of the entry to modify. |
| value | [in] Incoming value to add to the entry. |
Definition at line 4115 of file Tpetra_MultiVector_def.hpp.
|
inlineinherited |
Like the above replaceGlobalValue, but only enabled if T differs from impl_scalar_type.
This method only exists if its template parameter T and impl_scalar_type differ, and if it is syntactically possible to convert T to impl_scalar_type. This method is mainly useful for backwards compatibility, when the Scalar template parameter differs from impl_scalar_type. That is commonly only the case when Scalar is std::complex<U> for some type U.
This method affects the host memory version of the data. If device_type is a Kokkos device that has two memory spaces, and you want to modify the non-host version of the data, you must access the device View directly by calling getLocalView(). Please see modify(), sync(), and the discussion of DualView semantics elsewhere in the documentation. You are responsible for calling modify() and sync(), if needed; this method doesn't do that.
This method does not have an "atomic" option like sumIntoGlobalValue. This is deliberate. Replacement is not commutative, unlike += (modulo rounding error). Concurrent calls to replaceGlobalValue on different threads that modify the same entry/ies have undefined results. (It's not just that one thread might win; it's that the value might get messed up.)
| gblRow | [in] Global row index of the entry to modify. This must be a valid global row index on the calling process with respect to the MultiVector's Map. |
| col | [in] Column index of the entry to modify. |
| value | [in] Incoming value to add to the entry. |
Definition at line 933 of file Tpetra_MultiVector_decl.hpp.
|
inherited |
Update (+=) a value in host memory, using global row index.
Add the given value to the existing value at row gblRow (a global index) and column col. The column index is zero based.
This method affects the host memory version of the data. If device_type is a Kokkos device that has two memory spaces, and you want to modify the non-host version of the data, you must access the device View directly by calling getLocalView(). Please see modify(), sync(), and the discussion of DualView semantics elsewhere in the documentation. You are responsible for calling modify() and sync(), if needed; this method doesn't do that.
| gblRow | [in] Global row index of the entry to modify. This must be a valid global row index on the calling process with respect to the MultiVector's Map. |
| col | [in] Column index of the entry to modify. |
| value | [in] Incoming value to add to the entry. |
| atomic | [in] Whether to use an atomic update. If this class' execution space is not Kokkos::Serial, then this is true by default, else it is false by default. |
Definition at line 4139 of file Tpetra_MultiVector_def.hpp.
|
inlineinherited |
Like the above sumIntoGlobalValue, but only enabled if T differs from impl_scalar_type.
This method only exists if its template parameter T and impl_scalar_type differ, and if it is syntactically possible to convert T to impl_scalar_type. This method is mainly useful for backwards compatibility, when the Scalar template parameter differs from impl_scalar_type. That is commonly only the case when Scalar is std::complex<U> for some type U.
This method affects the host memory version of the data. If device_type is a Kokkos device that has two memory spaces, and you want to modify the non-host version of the data, you must access the device View directly by calling getLocalView(). Please see modify(), sync(), and the discussion of DualView semantics elsewhere in the documentation. You are responsible for calling modify() and sync(), if needed; this method doesn't do that.
| gblRow | [in] Global row index of the entry to modify. This must be a valid global row index on the calling process with respect to the MultiVector's Map. |
| col | [in] Column index of the entry to modify. |
| val | [in] Incoming value to add to the entry. |
| atomic | [in] Whether to use an atomic update. If this class' execution space is not Kokkos::Serial, then this is true by default, else it is false by default. |
Definition at line 998 of file Tpetra_MultiVector_decl.hpp.
|
inherited |
Replace value in host memory, using local (row) index.
Replace the current value at row lclRow (a local index) and column col with the given value. The column index is zero based.
This method affects the host memory version of the data. If device_type is a Kokkos device that has two memory spaces, and you want to modify the non-host version of the data, you must access the device View directly by calling getLocalView(). Please see modify(), sync(), and the discussion of DualView semantics elsewhere in the documentation. You are responsible for calling modify() and sync(), if needed; this method doesn't do that.
This method does not have an "atomic" option like sumIntoLocalValue. This is deliberate. Replacement is not commutative, unlike += (modulo rounding error). Concurrent calls to replaceLocalValue on different threads that modify the same entry/ies have undefined results. (It's not just that one thread might win; it's that the value might get messed up.)
| lclRow | [in] Local row index of the entry to modify. Must be a valid local index in this MultiVector's Map on the calling process. |
| col | [in] Column index of the entry to modify. |
| value | [in] Incoming value to add to the entry. |
Definition at line 4053 of file Tpetra_MultiVector_def.hpp.
|
inlineinherited |
Like the above replaceLocalValue, but only enabled if T differs from impl_scalar_type.
This method only exists if its template parameter T and impl_scalar_type differ, and if it is syntactically possible to convert T to impl_scalar_type. This method is mainly useful for backwards compatibility, when the Scalar template parameter differs from impl_scalar_type. That is commonly only the case when Scalar is std::complex<U> for some type U.
This method affects the host memory version of the data. If device_type is a Kokkos device that has two memory spaces, and you want to modify the non-host version of the data, you must access the device View directly by calling getLocalView(). Please see modify(), sync(), and the discussion of DualView semantics elsewhere in the documentation. You are responsible for calling modify() and sync(), if needed; this method doesn't do that.
This method does not have an "atomic" option like sumIntoLocalValue. This is deliberate. Replacement is not commutative, unlike += (modulo rounding error). Concurrent calls to replaceLocalValue on different threads that modify the same entry/ies have undefined results. (It's not just that one thread might win; it's that the value might get messed up.)
| lclRow | [in] Local row index of the entry to modify. Must be a valid local index in this MultiVector's Map on the calling process. |
| col | [in] Column index of the entry to modify. |
| val | [in] Incoming value to add to the entry. |
Definition at line 1073 of file Tpetra_MultiVector_decl.hpp.
|
inherited |
Update (+=) a value in host memory, using local row index.
Add the given value to the existing value at row localRow (a local index) and column col. The column index is zero based.
This method affects the host memory version of the data. If device_type is a Kokkos device that has two memory spaces, and you want to modify the non-host version of the data, you must access the device View directly by calling getLocalView(). Please see modify(), sync(), and the discussion of DualView semantics elsewhere in the documentation. You are responsible for calling modify() and sync(), if needed; this method doesn't do that.
| lclRow | [in] Local row index of the entry to modify. Must be a valid local index in this MultiVector's Map on the calling process. |
| col | [in] Column index of the entry to modify. |
| val | [in] Incoming value to add to the entry. |
| atomic | [in] Whether to use an atomic update. If this class' execution space is not Kokkos::Serial, then this is true by default, else it is false by default. |
Definition at line 4081 of file Tpetra_MultiVector_def.hpp.
|
inlineinherited |
Like the above sumIntoLocalValue, but only enabled if T differs from impl_scalar_type.
This method only exists if its template parameter T and impl_scalar_type differ, and if it is syntactically possible to convert T to impl_scalar_type. This method is mainly useful for backwards compatibility, when the Scalar template parameter differs from impl_scalar_type. That is commonly only the case when Scalar is std::complex<U> for some type U.
This method affects the host memory version of the data. If device_type is a Kokkos device that has two memory spaces, and you want to modify the non-host version of the data, you must access the device View directly by calling getLocalView(). Please see modify(), sync(), and the discussion of DualView semantics elsewhere in the documentation. You are responsible for calling modify() and sync(), if needed; this method doesn't do that.
| lclRow | [in] Local row index of the entry to modify. |
| col | [in] Column index of the entry to modify. |
| val | [in] Incoming value to add to the entry. |
| atomic | [in] Whether to use an atomic update. If this class' execution space is not Kokkos::Serial, then this is true by default, else it is false by default. |
Definition at line 1136 of file Tpetra_MultiVector_decl.hpp.
|
inherited |
Set all values in the multivector with the given value.
Definition at line 2319 of file Tpetra_MultiVector_def.hpp.
|
inlineinherited |
Set all values in the multivector with the given value.
This method only exists if its template parameter T and impl_scalar_type differ, and if it is syntactically possible to convert T to impl_scalar_type. This method is mainly useful for backwards compatibility, when the Scalar template parameter differs from impl_scalar_type. That is commonly only the case when Scalar is std::complex<U> for some type U.
Definition at line 1157 of file Tpetra_MultiVector_decl.hpp.
|
inherited |
Set all values in the multivector to pseudorandom numbers.
Definition at line 2255 of file Tpetra_MultiVector_def.hpp.
|
inherited |
Set all values in the multivector to pseudorandom numbers in the given range.
Definition at line 2272 of file Tpetra_MultiVector_def.hpp.
|
inherited |
Replace the underlying Map in place.
map->isCompatible (this->getMap ()). "Similar" means that the communicators have the same number of processes, though these need not be in the same order (have the same assignments of ranks) or represent the same communication contexts. It means the same thing as the MPI_SIMILAR return value of MPI_COMM_COMPARE. See MPI 3.0 Standard, Section 6.4.1.This method replaces this object's Map with the given Map. This relabels the rows of the multivector using the global IDs in the input Map. Thus, it implicitly applies a permutation, without actually moving data. If the new Map's communicator has more processes than the original Map's communicator, it "projects" the MultiVector onto the new Map by filling in missing rows with zeros. If the new Map's communicator has fewer processes than the original Map's communicator, the method "forgets about" any rows that do not exist in the new Map. (It mathematical terms, if one considers a MultiVector as a function from one vector space to another, this operation restricts the range.)
This method must always be called collectively on the communicator with the largest number of processes: either this object's current communicator (this->getMap()->getComm()), or the new Map's communicator (map->getComm()). If the new Map's communicator has fewer processes, then the new Map must be null on processes excluded from the original communicator, and the current Map must be nonnull on all processes. If the new Map has more processes, then it must be nonnull on all those processes, and the original Map must be null on those processes which are not in the new Map's communicator. (The latter case can only happen to a MultiVector to which a replaceMap() operation has happened before.)
this->getMap ()->getComm ()). We reserve the right to do checking in debug mode that requires this method to be called collectively in order not to deadlock.Definition at line 2368 of file Tpetra_MultiVector_def.hpp.
|
inherited |
Sum values of a locally replicated multivector across all processes.
Definition at line 4017 of file Tpetra_MultiVector_def.hpp.
|
inherited |
Return a MultiVector with copies of selected columns.
Definition at line 3006 of file Tpetra_MultiVector_def.hpp.
|
inherited |
Return a MultiVector with copies of selected columns.
Definition at line 2977 of file Tpetra_MultiVector_def.hpp.
|
inherited |
Return a const MultiVector with const views of selected columns.
Definition at line 3268 of file Tpetra_MultiVector_def.hpp.
|
inherited |
Return a const MultiVector with const views of selected columns.
Definition at line 3220 of file Tpetra_MultiVector_def.hpp.
|
inherited |
Return a MultiVector with views of selected columns.
Definition at line 3389 of file Tpetra_MultiVector_def.hpp.
|
inherited |
Return a MultiVector with views of selected columns.
Definition at line 3379 of file Tpetra_MultiVector_def.hpp.
|
inherited |
Return a Vector which is a const view of column j.
Definition at line 3447 of file Tpetra_MultiVector_def.hpp.
|
inherited |
Return a Vector which is a nonconst view of column j.
Definition at line 3457 of file Tpetra_MultiVector_def.hpp.
|
inherited |
Const view of the local values in a particular vector of this multivector.
Definition at line 2909 of file Tpetra_MultiVector_def.hpp.
|
inherited |
View of the local values in a particular vector of this multivector.
Definition at line 2941 of file Tpetra_MultiVector_def.hpp.
|
inherited |
Fill the given array with a copy of this multivector's local values.
| A | [out] View of the array to fill. We consider A as a matrix with column-major storage. |
| LDA | [in] Leading dimension of the matrix A. |
Definition at line 3467 of file Tpetra_MultiVector_def.hpp.
|
inherited |
Fill the given array with a copy of this multivector's local values.
| ArrayOfPtrs | [out] Array of arrays, one for each column of the multivector. On output, we fill ArrayOfPtrs[j] with the data for column j of this multivector. |
Definition at line 3542 of file Tpetra_MultiVector_def.hpp.
|
inherited |
Const persisting (1-D) view of this multivector's local values.
This method assumes that the columns of the multivector are stored contiguously. If not, this method throws std::runtime_error.
Definition at line 3599 of file Tpetra_MultiVector_def.hpp.
|
inherited |
Return const persisting pointers to values.
Definition at line 3676 of file Tpetra_MultiVector_def.hpp.
|
inherited |
Nonconst persisting (1-D) view of this multivector's local values.
This method assumes that the columns of the multivector are stored contiguously. If not, this method throws std::runtime_error.
Definition at line 3626 of file Tpetra_MultiVector_def.hpp.
|
inherited |
Return non-const persisting pointers to values.
Definition at line 3649 of file Tpetra_MultiVector_def.hpp.
|
inherited |
Clear "modified" flags on both host and device sides.
Definition at line 4192 of file Tpetra_MultiVector_def.hpp.
|
inlineinherited |
Update data on device or host only if data in the other space has been marked as modified.
If TargetDeviceType is the same as this MultiVector's device type, then copy data from host to device. Otherwise, copy data from device to host. In either case, only copy if the source of the copy has been modified.
This is a one-way synchronization only. If the target of the copy has been modified, this operation will discard those modifications. It will also reset both device and host modified flags.
Definition at line 1506 of file Tpetra_MultiVector_decl.hpp.
|
inherited |
Synchronize to Host.
Definition at line 4199 of file Tpetra_MultiVector_def.hpp.
|
inherited |
Synchronize to Device.
Definition at line 4206 of file Tpetra_MultiVector_def.hpp.
|
inlineinherited |
Whether this MultiVector needs synchronization to the given space.
Definition at line 1518 of file Tpetra_MultiVector_decl.hpp.
|
inherited |
Whether this MultiVector needs synchronization to the host.
Definition at line 4213 of file Tpetra_MultiVector_def.hpp.
|
inherited |
Whether this MultiVector needs synchronization to the device.
Definition at line 4220 of file Tpetra_MultiVector_def.hpp.
|
inlineinherited |
Mark data as modified on the given device TargetDeviceType.
If TargetDeviceType is the same as this MultiVector's device type, then mark the device's data as modified. Otherwise, mark the host's data as modified.
Definition at line 1534 of file Tpetra_MultiVector_decl.hpp.
|
inherited |
Mark data as modified on the device side.
Definition at line 4227 of file Tpetra_MultiVector_def.hpp.
|
inherited |
Mark data as modified on the host side.
Definition at line 4234 of file Tpetra_MultiVector_def.hpp.
|
inlineinherited |
Return a view of the local data on a specific device.
| TargetDeviceType | The Kokkos Device type whose data to return. |
Please don't be afraid of the if_c expression in the return value's type. That just tells the method what the return type should be: dual_view_type::t_dev if the TargetDeviceType template parameter matches this Tpetra object's device type, else dual_view_type::t_host.
For example, suppose you create a Tpetra::MultiVector for the Kokkos::Cuda device, like this:
If you want to get the CUDA device Kokkos::View, do this:
and if you want to get the host mirror of that View, do this:
Definition at line 1582 of file Tpetra_MultiVector_decl.hpp.
|
inherited |
A local Kokkos::View of host memory.
Definition at line 4248 of file Tpetra_MultiVector_def.hpp.
|
inherited |
A local Kokkos::View of device memory.
Definition at line 4241 of file Tpetra_MultiVector_def.hpp.
|
inherited |
Compute the dot product of each corresponding pair of vectors (columns) in A and B.
The "dot product" is the standard Euclidean inner product. If the type of entries of the vectors (impl_scalar_type) is complex, then A is transposed, not *this. For example, if x and y each have one column, then x.dot (y, dots) computes 
.
*this and A have the same number of columns (vectors). dots has at least as many entries as the number of columns in A.dots[j] == (this->getVector[j])->dot (* (A.getVector[j])) Definition at line 1932 of file Tpetra_MultiVector_def.hpp.
|
inlineinherited |
Compute the dot product of each corresponding pair of vectors (columns) in A and B.
| T | The output type of the dot products. |
This method only exists if dot_type and T are different types. For example, if impl_scalar_type and dot_type differ, then this method ensures backwards compatibility with the previous interface (that returned dot products as impl_scalar_type rather than as dot_type). The complicated enable_if expression just ensures that the method only exists if dot_type and T are different types; the method still returns void, as above.
Definition at line 1626 of file Tpetra_MultiVector_decl.hpp.
|
inlineinherited |
Like the above dot() overload, but for std::vector output.
Definition at line 1641 of file Tpetra_MultiVector_decl.hpp.
|
inherited |
Compute the dot product of each corresponding pair of vectors (columns) in A and B, storing the result in a device View.
The "dot product" is the standard Euclidean inner product. If the type of entries of the vectors (impl_scalar_type) is complex, then A is transposed, not *this. For example, if x and y each have one column, then x.dot (y, dots) computes .
| A | [in] MultiVector with which to dot *this. |
| dots | [out] Device View with getNumVectors() entries. |
this->getNumVectors () == A.getNumVectors () dots.extent (0) == A.getNumVectors ()dots(j) == (this->getVector[j])->dot (* (A.getVector[j])) Definition at line 1792 of file Tpetra_MultiVector_def.hpp.
|
inlineinherited |
Compute the dot product of each corresponding pair of vectors (columns) in A and B, storing the result in a device view.
| T | The output type of the dot products. |
This method only exists if dot_type and T are different types. For example, if Scalar and dot_type differ, then this method ensures backwards compatibility with the previous interface (that returned dot products as Scalar rather than as dot_type). The complicated enable_if expression just ensures that the method only exists if dot_type and T are different types; the method still returns void, as above.
Definition at line 1699 of file Tpetra_MultiVector_decl.hpp.
|
inherited |
Put element-wise absolute values of input Multi-vector in target: A = abs(this)
Definition at line 2721 of file Tpetra_MultiVector_def.hpp.
|
inherited |
Put element-wise reciprocal values of input Multi-vector in target, this(i,j) = 1/A(i,j).
Definition at line 2672 of file Tpetra_MultiVector_def.hpp.
|
inherited |
Scale in place: this = alpha*this.
Replace this MultiVector with alpha times this MultiVector. This method will always multiply, even if alpha is zero. That means, for example, that if *this contains NaN entries before calling this method, the NaN entries will remain after this method finishes.
Definition at line 2465 of file Tpetra_MultiVector_def.hpp.
|
inherited |
Scale each column in place: this[j] = alpha[j]*this[j].
Replace each column j of this MultiVector with alpha[j] times the current column j of this MultiVector. This method will always multiply, even if all the entries of alpha are zero. That means, for example, that if *this contains NaN entries before calling this method, the NaN entries will remain after this method finishes.
Definition at line 2518 of file Tpetra_MultiVector_def.hpp.
|
inherited |
Scale each column in place: this[j] = alpha[j]*this[j].
Replace each column j of this MultiVector with alpha[j] times the current column j of this MultiVector. This method will always multiply, even if all the entries of alpha are zero. That means, for example, that if *this contains NaN entries before calling this method, the NaN entries will remain after this method finishes.
Definition at line 2542 of file Tpetra_MultiVector_def.hpp.
|
inherited |
Scale in place: this = alpha * A.
Replace this MultiVector with scaled values of A. This method will always multiply, even if alpha is zero. That means, for example, that if *this contains NaN entries before calling this method, the NaN entries will remain after this method finishes. It is legal for the input A to alias this MultiVector.
Definition at line 2613 of file Tpetra_MultiVector_def.hpp.
|
inherited |
Update: this = beta*this + alpha*A.
Update this MultiVector with scaled values of A. If beta is zero, overwrite *this unconditionally, even if it contains NaN entries. It is legal for the input A to alias this MultiVector.
Definition at line 2770 of file Tpetra_MultiVector_def.hpp.
|
inherited |
Update: this = gamma*this + alpha*A + beta*B.
Update this MultiVector with scaled values of A and B. If gamma is zero, overwrite *this unconditionally, even if it contains NaN entries. It is legal for the inputs A or B to alias this MultiVector.
Definition at line 2832 of file Tpetra_MultiVector_def.hpp.
|
inherited |
Compute the one-norm of each vector (column), storing the result in a host view.
| norms | [out] Host View with getNumVectors() entries. |
norms.extent (0) == this->getNumVectors () norms(j) == (this->getVector[j])->norm1 (* (A.getVector[j]))The one-norm of a vector is the sum of the magnitudes of the vector's entries. On exit, norms(j) is the one-norm of column j of this MultiVector.
Definition at line 2113 of file Tpetra_MultiVector_def.hpp.
|
inlineinherited |
Compute the one-norm of each vector (column), storing the result in a device view.
| T | The output type of the dot products. |
See the above norm1() method for documentation.
This method only exists if mag_type and T are different types. For example, if Teuchos::ScalarTraits<Scalar>::magnitudeType and mag_type differ, then this method ensures backwards compatibility with the previous interface (that returned norms products as Teuchos::ScalarTraits<Scalar>::magnitudeType rather than as mag_type). The complicated enable_if expression just ensures that the method only exists if mag_type and T are different types; the method still returns void, as above.
Definition at line 1831 of file Tpetra_MultiVector_decl.hpp.
|
inherited |
Compute the one-norm of each vector (column).
See the uppermost norm1() method above for documentation.
Definition at line 2098 of file Tpetra_MultiVector_def.hpp.
|
inlineinherited |
Compute the one-norm of each vector (column).
| T | The output type of the norms. |
See the uppermost norm1() method above for documentation.
This method only exists if mag_type and T are different types. For example, if Teuchos::ScalarTraits<Scalar>::magnitudeType and mag_type differ, then this method ensures backwards compatibility with the previous interface (that returned norms as Teuchos::ScalarTraits<Scalar>::magnitudeType rather than as mag_type). The complicated enable_if expression just ensures that the method only exists if mag_type and T are different types; the method still returns void, as above.
Definition at line 1866 of file Tpetra_MultiVector_decl.hpp.
|
inherited |
Compute the two-norm of each vector (column), storing the result in a host View.
| norms | [out] Host View with getNumVectors() entries. |
norms.extent (0) == this->getNumVectors () norms(j) == (this->getVector[j])->dot (* (A.getVector[j]))The two-norm of a vector is the standard Euclidean norm, the square root of the sum of squares of the magnitudes of the vector's entries. On exit, norms(k) is the two-norm of column k of this MultiVector.
Definition at line 1994 of file Tpetra_MultiVector_def.hpp.
|
inlineinherited |
Compute the two-norm of each vector (column), storing the result in a device view.
See the above norm2() method for documentation.
This method only exists if mag_type and T are different types. For example, if Teuchos::ScalarTraits<Scalar>::magnitudeType and mag_type differ, then this method ensures backwards compatibility with the previous interface (that returned norms as Teuchos::ScalarTraits<Scalar>::magnitudeType rather than as mag_type). The complicated enable_if expression just ensures that the method only exists if mag_type and T are different types; the method still returns void, as above.
Definition at line 1921 of file Tpetra_MultiVector_decl.hpp.
|
inherited |
Compute the two-norm of each vector (column).
See the uppermost norm2() method above for documentation.
Definition at line 1979 of file Tpetra_MultiVector_def.hpp.
|
inlineinherited |
Compute the two-norm of each vector (column).
| T | The output type of the norms. |
See the uppermost norm2() method above for documentation.
This method only exists if mag_type and T are different types. For example, if Teuchos::ScalarTraits<Scalar>::magnitudeType and mag_type differ, then this method ensures backwards compatibility with the previous interface (that returned norms products as Teuchos::ScalarTraits<Scalar>::magnitudeType rather than as mag_type). The complicated enable_if expression just ensures that the method only exists if mag_type and T are different types; the method still returns void, as above.
Definition at line 1956 of file Tpetra_MultiVector_decl.hpp.
|
inherited |
Compute the infinity-norm of each vector (column), storing the result in a host View.
The infinity-norm of a vector is the maximum of the magnitudes of the vector's entries. On exit, norms(j) is the infinity-norm of column j of this MultiVector.
Definition at line 2137 of file Tpetra_MultiVector_def.hpp.
|
inlineinherited |
Compute the infinity-norm of each vector (column), storing the result in a device view.
See the above normInf() method for documentation.
This method only exists if mag_type and T are different types. For example, if Teuchos::ScalarTraits<Scalar>::magnitudeType and mag_type differ, then this method ensures backwards compatibility with the previous interface (that returned norms as Teuchos::ScalarTraits<Scalar>::magnitudeType rather than as mag_type). The complicated enable_if expression just ensures that the method only exists if mag_type and T are different types; the method still returns void, as above.
Definition at line 2004 of file Tpetra_MultiVector_decl.hpp.
|
inherited |
Compute the infinity-norm of each vector (column), storing the result in a Teuchos::ArrayView.
See the uppermost normInf() method above for documentation.
Definition at line 2122 of file Tpetra_MultiVector_def.hpp.
|
inlineinherited |
Compute the infinity-norm of each vector (column), storing the result in a Teuchos::ArrayView.
| T | The output type of the norms. |
See the uppermost normInf() method above for documentation.
This method only exists if mag_type and T are different types. For example, if Teuchos::ScalarTraits<Scalar>::magnitudeType and mag_type differ, then this method ensures backwards compatibility with the previous interface (that returned norms products as Teuchos::ScalarTraits<Scalar>::magnitudeType rather than as mag_type). The complicated enable_if expression just ensures that the method only exists if mag_type and T are different types; the method still returns void, as above.
Definition at line 2041 of file Tpetra_MultiVector_decl.hpp.
|
inherited |
Compute mean (average) value of each column.
The outcome of this routine is undefined for non-floating point scalar types (e.g., int).
Definition at line 2146 of file Tpetra_MultiVector_def.hpp.
|
inherited |
Matrix-matrix multiplication: this = beta*this + alpha*op(A)*op(B).
If beta is zero, overwrite *this unconditionally, even if it contains NaN entries. This imitates the semantics of analogous BLAS routines like DGEMM.
Definition at line 3707 of file Tpetra_MultiVector_def.hpp.
|
inherited |
Multiply a Vector A elementwise by a MultiVector B.
Compute this = scalarThis * this + scalarAB * B @ A where </tt> denotes element-wise multiplication. In pseudocode, if C denotes *this MultiVector:
for all rows i and columns j of C.
B must have the same dimensions as *this, while A must have the same number of rows but a single column.
We do not require that A, B, and *this have compatible Maps, as long as the number of rows in A, B, and *this on each process is the same. For example, one or more of these vectors might have a locally replicated Map, or a Map with a local communicator (MPI_COMM_SELF). This case may occur in block relaxation algorithms when applying a diagonal scaling.
Definition at line 3951 of file Tpetra_MultiVector_def.hpp.
|
inherited |
Number of columns in the multivector.
Definition at line 1728 of file Tpetra_MultiVector_def.hpp.
|
inherited |
Local number of rows on the calling process.
Definition at line 807 of file Tpetra_MultiVector_def.hpp.
|
inherited |
Global number of rows in the multivector.
Definition at line 819 of file Tpetra_MultiVector_def.hpp.
|
inherited |
Stride between columns in the multivector.
This is only meaningful if isConstantStride() returns true.
Definition at line 831 of file Tpetra_MultiVector_def.hpp.
|
inherited |
Whether this multivector has constant stride between columns.
Definition at line 800 of file Tpetra_MultiVector_def.hpp.
|
virtualinherited |
Remove processes owning zero rows from the Map and their communicator.
| newMap | [in] This must be the result of calling the removeEmptyProcesses() method on the row Map. If it is not, this method's behavior is undefined. This pointer will be null on excluded processes. |
Definition at line 4470 of file Tpetra_MultiVector_def.hpp.
|
virtualinherited |
Remove processes which contain no entries in this object's Map.
On input, this object is distributed over the Map returned by getMap() (the "original Map," with its communicator, the "original communicator"). The input newMap of this method must be the same as the result of calling getMap()->removeEmptyProcesses(). On processes in the original communicator which contain zero entries ("excluded processes," as opposed to "included processes"), the input newMap must be Teuchos::null (which is what getMap()->removeEmptyProcesses() returns anyway).
On included processes, reassign this object's Map (that would be returned by getMap()) to the input newMap, and do any work that needs to be done to restore correct semantics. On excluded processes, free any data that needs freeing, and do any other work that needs to be done to restore correct semantics.
This method has collective semantics over the original communicator. On exit, the only method of this object which is safe to call on excluded processes is the destructor. This implies that subclasses' destructors must not contain communication operations.
|
inlineinherited |
Set whether this has copy (copyOrView = Teuchos::Copy) or view (copyOrView = Teuchos::View) semantics.
Definition at line 2248 of file Tpetra_MultiVector_decl.hpp.
|
inlineinherited |
Get whether this has copy (copyOrView = Teuchos::Copy) or view (copyOrView = Teuchos::View) semantics.
Definition at line 2265 of file Tpetra_MultiVector_decl.hpp.
|
inherited |
Copy the contents of src into *this (deep copy).
| src | [in] Source MultiVector (input of the deep copy). |
! src.getMap ().is_null () && ! this->getMap ().is_null () src.getMap ()->isCompatible (* (this->getMap ()) src or *this remain valid.*this, or otherwise change its dimensions. This is not an assignment operator; it does not change anything in *this other than the contents of storage. Definition at line 4478 of file Tpetra_MultiVector_def.hpp.
|
inherited |
| src | [in] MultiVector |
! vec.getMap ().is_null () && ! this->getMap ().is_null () vec.getMap ()->isCompatible (* (this->getMap ()) src or *this remain valid. Definition at line 4528 of file Tpetra_MultiVector_def.hpp.
|
protectedinherited |
Implementation of description() for this class, and its subclass Vector.
| className | [in] Name of the class calling this method: Either "Tpetra::MultiVector" or "Tpetra::Vector" (no quotes in the string, in either case). |
Definition at line 4255 of file Tpetra_MultiVector_def.hpp.
|
protectedinherited |
Print the calling process' verbose describe() information to the returned string.
This is an implementation detail of describe().
| vl | [in] Verbosity level with which to print. |
Definition at line 4293 of file Tpetra_MultiVector_def.hpp.
|
protectedinherited |
Implementation of describe() for this class, and its subclass Vector.
| out | [out] Output stream to which to write. Only Process 0 in this object's communicator may write to the output stream. |
| className | [in] Name of the class calling this method. |
| verbLevel | [in] Verbosity level. This also controls whether this method does any communication. At verbosity levels higher (greater) than Teuchos::VERB_LOW, this method behaves as a collective over the object's communicator. |
Definition at line 4388 of file Tpetra_MultiVector_def.hpp.
|
protectedinherited |
Persisting view of j-th column in the given ArrayRCP.
This method considers isConstantStride(). The ArrayRCP may correspond either to a compute buffer or a host view.
Definition at line 4167 of file Tpetra_MultiVector_def.hpp.
|
protectedinherited |
"Original" number of rows in the (local) data.
Definition at line 3020 of file Tpetra_MultiVector_def.hpp.
|
protectedinherited |
"Original" number of columns in the (local) data.
Definition at line 3027 of file Tpetra_MultiVector_def.hpp.
|
protectedvirtualinherited |
Whether data redistribution between sourceObj and this object is legal.
This method is called in DistObject::doTransfer() to check whether data redistribution between the two objects is legal.
Implements Tpetra::DistObject< Scalar, LocalOrdinal, GlobalOrdinal, Node >.
Definition at line 839 of file Tpetra_MultiVector_def.hpp.
|
protectedvirtualinherited |
Number of packets to send per LID.
Reimplemented from Tpetra::DistObject< Scalar, LocalOrdinal, GlobalOrdinal, Node >.
Definition at line 863 of file Tpetra_MultiVector_def.hpp.
|
protectedvirtualinherited |
Perform copies and permutations that are local to the calling (MPI) process.
Subclasses must reimplement this function. Its default implementation does nothing. Note that the <t>target object of the Export or Import, namely *this, packs the source object's data.
permuteToLIDs.need_sync_host(), permuteToLIDs.need_sync_device(), permuteFromLIDs.need_sync_host(), and permuteFromLIDs.need_sync_device() are all false.| source | [in] On entry, the source object of the Export or Import operation. |
| numSameIDs | [in] The number of elements that are the same on the source and target objects. These elements live on the same process in both the source and target objects. |
| permuteToLIDs | [in] List of the elements that are permuted. They are listed by their local index (LID) in the destination object. |
| permuteFromLIDs | [in] List of the elements that are permuted. They are listed by their local index (LID) in the source object. |
|
protectedvirtualinherited |
Pack data and metadata for communication (sends).
Subclasses must reimplement this function. Its default implementation does nothing. Note that the <t>target object of the Export or Import, namely *this, packs the source object's data.
exportLIDs.need_sync_host () and exportLIDs.need_sync_device() are both false.| source | [in] Source object for the redistribution. |
| exportLIDs | [in] List of the entries (as local IDs in the source object) that Tpetra will send to other processes. |
| exports | [out] On exit, the packed data to send. Implementations must reallocate this as needed (prefer reusing the existing allocation if possible), and may modify and/or sync this wherever they like. |
| numPacketsPerLID | [out] On exit, the implementation of this method must do one of two things: either set numPacketsPerLID[i] to the number of packets to be packed for exportLIDs[i] and set constantNumPackets to zero, or set constantNumPackets to a nonzero value. If the latter, the implementation must not modify the entries of numPacketsPerLID. If the former, the implementation may sync numPacketsPerLID this wherever it likes, either to host or to device. The allocation belongs to DistObject, not to subclasses; don't be tempted to change this to pass by reference. |
| constantNumPackets | [out] On exit, 0 if the number of packets per LID could differ, else (if nonzero) the number of packets per LID (which must be constant). |
| distor | [in] The Distributor object we are using. Most implementations will not use this. |
|
protectedvirtualinherited |
Perform any unpacking and combining after communication.
Subclasses must reimplement this function. Its default implementation does nothing. Note that the <t>target object of the Export or Import, namely *this, unpacks the received data into itself, possibly modifying its entries.
importLIDs.need_sync_host () and importLIDs.need_sync_device() are both false.| importLIDs | [in] List of the entries (as LIDs in the destination object) we received from other processes. |
| imports | [in/out] On input: Buffer of received data to unpack. DistObject promises nothing about where this is sync'd. Implementations may sync this wherever they like, either to host or to device. The allocation belongs to DistObject, not to subclasses; don't be tempted to change this to pass by reference. |
| numPacketsPerLID | [in/out] On input: If constantNumPackets is zero, then numPacketsPerLID[i] contains the number of packets imported for importLIDs[i]. DistObject promises nothing about where this is sync'd. Implementations may sync this wherever they like, either to host or to device. The allocation belongs to DistObject, not to subclasses; don't be tempted to change this to pass by reference. |
| constantNumPackets | [in] If nonzero, then the number of packets per LID is the same for all entries ("constant") and constantNumPackets is that number. If zero, then numPacketsPerLID[i] is the number of packets to unpack for LID importLIDs[i]. |
| distor | [in] The Distributor object we are using. Most implementations will not use this. |
| combineMode | [in] The CombineMode to use when combining the imported entries with existing entries. |
|
inherited |
Import data into this object using an Import object ("forward mode").
The input DistObject is always the source of the data redistribution operation, and the *this object is always the target.
If you don't know the difference between forward and reverse mode, then you probably want forward mode. Use this method with your precomputed Import object if you want to do an Import, else use doExport() with a precomputed Export object.
"Restricted Mode" does two things:
*this, in a "locallyFitted" sense. This cannot be used if (2) is not true, OR there are permutes. The "source" maps still need to match.
| source | [in] The "source" object for redistribution. |
| importer | [in] Precomputed data redistribution plan. Its source Map must be the same as the input DistObject's Map, and its target Map must be the same as this->getMap(). |
| CM | [in] How to combine incoming data with the same global index. |
|
inherited |
Import data into this object using an Export object ("reverse mode").
The input DistObject is always the source of the data redistribution operation, and the *this object is always the target.
If you don't know the difference between forward and reverse mode, then you probably want forward mode. Use the version of doImport() that takes a precomputed Import object in that case.
"Restricted Mode" does two things:
*this, in a "locallyFitted" sense. This cannot be used if (2) is not true, OR there are permutes. The "source" maps still need to match.
| source | [in] The "source" object for redistribution. |
| exporter | [in] Precomputed data redistribution plan. Its target Map must be the same as the input DistObject's Map, and its source Map must be the same as this->getMap(). (Note the difference from forward mode.) |
| CM | [in] How to combine incoming data with the same global index. |
|
inherited |
Export data into this object using an Export object ("forward mode").
The input DistObject is always the source of the data redistribution operation, and the *this object is always the target.
If you don't know the difference between forward and reverse mode, then you probably want forward mode. Use this method with your precomputed Export object if you want to do an Export, else use doImport() with a precomputed Import object.
"Restricted Mode" does two things:
*this, in a "locallyFitted" sense. This cannot be used if (2) is not true, OR there are permutes. The "source" maps still need to match.
| source | [in] The "source" object for redistribution. |
| exporter | [in] Precomputed data redistribution plan. Its source Map must be the same as the input DistObject's Map, and its target Map must be the same as this->getMap(). |
| CM | [in] How to combine incoming data with the same global index. |
|
inherited |
Export data into this object using an Import object ("reverse mode").
The input DistObject is always the source of the data redistribution operation, and the *this object is always the target.
If you don't know the difference between forward and reverse mode, then you probably want forward mode. Use the version of doExport() that takes a precomputed Export object in that case.
"Restricted Mode" does two things:
*this, in a "locallyFitted" sense. This cannot be used if (2) is not true, OR there are permutes. The "source" maps still need to match.
| source | [in] The "source" object for redistribution. |
| importer | [in] Precomputed data redistribution plan. Its target Map must be the same as the input DistObject's Map, and its source Map must be the same as this->getMap(). (Note the difference from forward mode.) |
| CM | [in] How to combine incoming data with the same global index. |
|
inherited |
Whether this is a globally distributed object.
For a definition of "globally distributed" (and its opposite, "locally replicated"), see the documentation of Map's isDistributed() method.
|
inlinevirtualinherited |
The Map describing the parallel distribution of this object.
Note that some Tpetra objects might be distributed using multiple Map objects. For example, CrsMatrix has both a row Map and a column Map. It is up to the subclass to decide which Map to use when invoking the DistObject constructor.
Definition at line 541 of file Tpetra_DistObject_decl.hpp.
|
inherited |
Print this object to the given output stream.
We generally assume that all MPI processes can print to the given stream.
|
protectedvirtualinherited |
Redistribute data across (MPI) processes.
| src | [in] The source object, to redistribute into the target object, which is *this object. |
| transfer | [in] The Export or Import object representing the communication pattern. (Details::Transfer is the common base class of these two objects.) |
| modeString | [in] Human-readable string, for verbose debugging output and error output, explaining what function called this method. Example: "doImport (forward)", "doExport (reverse)". |
| revOp | [in] Whether to do a forward or reverse mode redistribution. |
| CM | [in] The combine mode that describes how to combine values that map to the same global ID on the same process. |
|
protectedvirtualinherited |
Reallocate numExportPacketsPerLID_ and/or numImportPacketsPerLID_, if necessary.
| numExportLIDs | [in] Number of entries in the exportLIDs input array argument of doTransfer(). |
| numImportLIDs | [in] Number of entries in the remoteLIDs input array argument of doTransfer(). |
|
protectedvirtualinherited |
Implementation detail of doTransfer.
LID DualViews come from the Transfer object given to doTransfer. They are always sync'd on both host and device. Users must never attempt to modify or sync them.
|
protectedinherited |
Reallocate imports_ if needed.
This unfortunately must be declared protected, for the same reason that imports_ is declared protected.
| newSize | [in] New size of imports_. |
| verbose | [in] Whether to print verbose debugging output to stderr on every (MPI) process in the communicator. |
| prefix | [in] If verbose is true, then this is a nonnull prefix to print at the beginning of each line of verbose debugging output. Otherwise, not used. |
We don't need a "reallocExportsIfNeeded" method, because exports_ always gets passed into packAndPrepare() by nonconst reference. Thus, that method can resize the DualView without needing to call other DistObject methods.
|
related |
Nonmember MultiVector "constructor": Create a MultiVector from a given Map.
| map | [in] Map describing the distribution of rows of the resulting MultiVector. |
| numVectors | [in] Number of columns of the resulting MultiVector. |
|
related |
Return a deep copy of the given Vector.
|
related |
Copy the contents of the MultiVector src into dst.
src must be compatible with the Map of dst. Copy the contents of the MultiVector src into the MultiVector dst. ("Copy the contents" means the same thing as "deep
copy.") The two MultiVectors need not necessarily have the same template parameters, but the assignment of their entries must make sense. Furthermore, their Maps must be compatible, that is, the MultiVectors' local dimensions must be the same on all processes.
This method must always be called as a collective operation on all processes over which the multivector is distributed. This is because the method reserves the right to check for compatibility of the two Maps, at least in debug mode, and throw if they are not compatible.
|
related |
Return a deep copy of the given MultiVector.
|
staticprotectedinherited |
Whether sumIntoLocalValue and sumIntoGlobalValue should use atomic updates by default.
Definition at line 858 of file Tpetra_MultiVector_decl.hpp.
|
mutableprotectedinherited |
The Kokkos::DualView containing the MultiVector's data.
This has to be declared mutable, so that get1dView() can retain its current const marking, even though it has always implied a device->host synchronization. Lesson to the reader: Use const sparingly!
Definition at line 2315 of file Tpetra_MultiVector_decl.hpp.
|
mutableprotectedinherited |
The "original view" of the MultiVector's data.
Methods like offsetView() return a view of a contiguous subset of rows. At some point, we might like to get all of the rows back, by taking another view of a superset of rows. For example, we might like to get a column Map view of a (domain Map view of a (column Map MultiVector)). Tpetra's implementation of Gauss-Seidel and SOR in CrsMatrix relies on this functionality. However, Kokkos (rightfully) forbids us from taking a superset of rows of the current view.
We deal with this at the Tpetra level by keeping around the original view of all the rows (and columns), which is origView_. Methods like offsetView() then use origView_, not view_, to make the subview for the returned MultiVector. Furthermore, offsetView() can do error checking by getting the original number of rows from origView_.
This may pose some problems for offsetView if it is given an offset other than zero, but that case is hardly exercised, so I am not going to worry about it for now.
Note that the "original" view isn't always original. It always has the original number of rows. However, some special cases of constructors that take a whichVectors argument, when whichVectors.size() is 1, may point origView_ to the column to view. Those constructors do this so that the resulting MultiVector has constant stride. This special case does not affect correctness of offsetView and related methods.
Definition at line 2346 of file Tpetra_MultiVector_decl.hpp.
|
protectedinherited |
Indices of columns this multivector is viewing.
If this array has nonzero size, then this multivector is a view of another multivector (the "original" multivector). In that case, whichVectors_ contains the indices of the columns of the original multivector. Furthermore, isConstantStride() returns false in this case.
If this array has zero size, then this multivector is not a view of any other multivector. Furthermore, the stride between columns of this multivector is a constant: thus, isConstantStride() returns true.
Definition at line 2360 of file Tpetra_MultiVector_decl.hpp.
|
protectedinherited |
The Map over which this object is distributed.
Definition at line 1140 of file Tpetra_DistObject_decl.hpp.
|
protectedinherited |
Buffer into which packed data are imported (received from other processes).
Unfortunately, I had to declare these protected, because CrsMatrix uses them at one point. Please, nobody else use them.
Definition at line 1149 of file Tpetra_DistObject_decl.hpp.
|
protectedinherited |
Number of packets to receive for each receive operation.
This array is used in Distributor::doPosts() (and doReversePosts()) when starting the ireceive operation.
This may be ignored in doTransfer() if constantNumPackets is nonzero, indicating a constant number of packets per LID. (For example, MultiVector sets the constantNumPackets output argument of packAndPrepare() to the number of columns in the multivector.)
Unfortunately, I had to declare this protected, because CrsMatrix uses it at one point. Please, nobody else use it.
Definition at line 1187 of file Tpetra_DistObject_decl.hpp.
|
protectedinherited |
Buffer from which packed data are exported (sent to other processes).
Unfortunately, I had to declare this protected, because CrsMatrix uses it at one point. Please, nobody else use it.
Definition at line 1194 of file Tpetra_DistObject_decl.hpp.
|
protectedinherited |
Number of packets to send for each send operation.
This array is used in Distributor::doPosts() (and doReversePosts()) for preparing for the send operation.
This may be ignored in doTransfer() if constantNumPackets is nonzero, indicating a constant number of packets per LID. (For example, MultiVector sets the constantNumPackets output argument of packAndPrepare() to the number of columns in the multivector.)
Unfortunately, I had to declare this protected, because CrsMatrix uses them at one point. Please, nobody else use it.
Definition at line 1209 of file Tpetra_DistObject_decl.hpp.
1.8.5