Class for Matrices with few columns that consists of Vectors. More...
#include <IpMultiVectorMatrix.hpp>
Public Member Functions | |
SmartPtr< MultiVectorMatrix > | MakeNewMultiVectorMatrix () const |
Create a new MultiVectorMatrix from same MatrixSpace. | |
SmartPtr< const Vector > | GetVector (Index i) const |
Get a Vector in a particular column as a const Vector. | |
SmartPtr< Vector > | GetVectorNonConst (Index i) |
Get a Vector in a particular column as a non-const Vector. | |
void | ScaleRows (const Vector &scal_vec) |
Method for scaling the rows of the matrix, using the ElementWiseMultiply method for each column vector. | |
void | ScaleColumns (const Vector &scal_vec) |
Method for scaling the columns of the matrix, using the Scal method for each column vector. | |
void | AddOneMultiVectorMatrix (Number a, const MultiVectorMatrix &mv1, Number c) |
Adding another MultiVectorMatrix, using the AddOneVector methods for the individual column vectors. | |
void | AddRightMultMatrix (Number a, const MultiVectorMatrix &U, const Matrix &C, Number b) |
Multiplying a Matrix C (for now assumed to be a DenseGenMatrix) from the right to a MultiVectorMatrix U and adding the result to this MultiVectorMatrix V. | |
void | FillWithNewVectors () |
Method for initializing all Vectors with new (uninitialized) Vectors. | |
void | LRMultVector (Number alpha, const Vector &x, Number beta, Vector &y) const |
Method for adding the low-rank update matrix corresponding to this matrix to a vector. | |
SmartPtr< const VectorSpace > | ColVectorSpace () const |
Vector space for the columns. | |
SmartPtr< const MultiVectorMatrixSpace > | MultiVectorMatrixOwnerSpace () const |
Return the MultiVectorMatrixSpace. | |
Constructors / Destructors | |
MultiVectorMatrix (const MultiVectorMatrixSpace *owner_space) | |
Constructor, taking the owner_space. | |
~MultiVectorMatrix () | |
Destructor. | |
void | SetVector (Index i, const Vector &vec) |
Set a particular Vector at a given column position, replacing another vector if there has been one. | |
void | SetVectorNonConst (Index i, Vector &vec) |
Set a particular Vector at a given column position, replacing another vector if there has been one. | |
Protected Member Functions | |
Overloaded methods from Matrix base class | |
virtual void | MultVectorImpl (Number alpha, const Vector &x, Number beta, Vector &y) const |
Method for determining if all stored numbers are valid (i.e., no Inf or Nan). | |
virtual void | TransMultVectorImpl (Number alpha, const Vector &x, Number beta, Vector &y) const |
Method for determining if all stored numbers are valid (i.e., no Inf or Nan). | |
virtual bool | HasValidNumbersImpl () const |
Method for determining if all stored numbers are valid (i.e., no Inf or Nan). | |
virtual void | ComputeRowAMaxImpl (Vector &rows_norms, bool init) const |
Method for determining if all stored numbers are valid (i.e., no Inf or Nan). | |
virtual void | ComputeColAMaxImpl (Vector &cols_norms, bool init) const |
Method for determining if all stored numbers are valid (i.e., no Inf or Nan). | |
virtual void | PrintImpl (const Journalist &jnlst, EJournalLevel level, EJournalCategory category, const std::string &name, Index indent, const std::string &prefix) const |
Method for determining if all stored numbers are valid (i.e., no Inf or Nan). | |
Private Member Functions | |
Default Compiler Generated Methods | |
MultiVectorMatrix () | |
Default Constructor. | |
MultiVectorMatrix (const MultiVectorMatrix &) | |
Copy Constructor. | |
void | operator= (const MultiVectorMatrix &) |
Overloaded Equals Operator. | |
const Vector * | ConstVec (Index i) const |
Method for accessing the internal Vectors internally. | |
Vector * | Vec (Index i) |
Method for accessing the internal Vectors internally. | |
Private Attributes | |
const MultiVectorMatrixSpace * | owner_space_ |
std::vector< SmartPtr< const Vector > > | const_vecs_ |
space for storing the const Vector's | |
std::vector< SmartPtr< Vector > > | non_const_vecs_ |
space for storing the non-const Vector's |
Class for Matrices with few columns that consists of Vectors.
Those matrices are for example useful in the implementation of limited memory quasi-Newton methods.
Definition at line 25 of file IpMultiVectorMatrix.hpp.
Ipopt::MultiVectorMatrix::MultiVectorMatrix | ( | const MultiVectorMatrixSpace * | owner_space | ) |
Constructor, taking the owner_space.
Ipopt::MultiVectorMatrix::~MultiVectorMatrix | ( | ) | [inline] |
Destructor.
Definition at line 225 of file IpMultiVectorMatrix.hpp.
Ipopt::MultiVectorMatrix::MultiVectorMatrix | ( | ) | [private] |
Default Constructor.
Ipopt::MultiVectorMatrix::MultiVectorMatrix | ( | const MultiVectorMatrix & | ) | [private] |
Copy Constructor.
SmartPtr< MultiVectorMatrix > Ipopt::MultiVectorMatrix::MakeNewMultiVectorMatrix | ( | ) | const [inline] |
Create a new MultiVectorMatrix from same MatrixSpace.
Definition at line 229 of file IpMultiVectorMatrix.hpp.
Get a Vector in a particular column as a const Vector.
Definition at line 56 of file IpMultiVectorMatrix.hpp.
Get a Vector in a particular column as a non-const Vector.
This is fail if the column has currently only a non-const Vector stored.
Definition at line 64 of file IpMultiVectorMatrix.hpp.
void Ipopt::MultiVectorMatrix::ScaleRows | ( | const Vector & | scal_vec | ) |
Method for scaling the rows of the matrix, using the ElementWiseMultiply method for each column vector.
void Ipopt::MultiVectorMatrix::ScaleColumns | ( | const Vector & | scal_vec | ) |
Method for scaling the columns of the matrix, using the Scal method for each column vector.
void Ipopt::MultiVectorMatrix::AddOneMultiVectorMatrix | ( | Number | a, | |
const MultiVectorMatrix & | mv1, | |||
Number | c | |||
) |
Adding another MultiVectorMatrix, using the AddOneVector methods for the individual column vectors.
void Ipopt::MultiVectorMatrix::AddRightMultMatrix | ( | Number | a, | |
const MultiVectorMatrix & | U, | |||
const Matrix & | C, | |||
Number | b | |||
) |
Multiplying a Matrix C (for now assumed to be a DenseGenMatrix) from the right to a MultiVectorMatrix U and adding the result to this MultiVectorMatrix V.
V = a * U * C + b * V.
void Ipopt::MultiVectorMatrix::FillWithNewVectors | ( | ) |
Method for initializing all Vectors with new (uninitialized) Vectors.
void Ipopt::MultiVectorMatrix::LRMultVector | ( | Number | alpha, | |
const Vector & | x, | |||
Number | beta, | |||
Vector & | y | |||
) | const |
Method for adding the low-rank update matrix corresponding to this matrix to a vector.
If V is this MultiVectorMatrix, the operation is y = beta*y + alpha*V*V^T*x.
SmartPtr< const VectorSpace > Ipopt::MultiVectorMatrix::ColVectorSpace | ( | ) | const [inline] |
Vector space for the columns.
Definition at line 235 of file IpMultiVectorMatrix.hpp.
SmartPtr< const MultiVectorMatrixSpace > Ipopt::MultiVectorMatrix::MultiVectorMatrixOwnerSpace | ( | ) | const [inline] |
Return the MultiVectorMatrixSpace.
Definition at line 242 of file IpMultiVectorMatrix.hpp.
virtual void Ipopt::MultiVectorMatrix::MultVectorImpl | ( | Number | alpha, | |
const Vector & | x, | |||
Number | beta, | |||
Vector & | y | |||
) | const [protected, virtual] |
Method for determining if all stored numbers are valid (i.e., no Inf or Nan).
Implements Ipopt::Matrix.
virtual void Ipopt::MultiVectorMatrix::TransMultVectorImpl | ( | Number | alpha, | |
const Vector & | x, | |||
Number | beta, | |||
Vector & | y | |||
) | const [protected, virtual] |
Method for determining if all stored numbers are valid (i.e., no Inf or Nan).
Implements Ipopt::Matrix.
virtual bool Ipopt::MultiVectorMatrix::HasValidNumbersImpl | ( | ) | const [protected, virtual] |
Method for determining if all stored numbers are valid (i.e., no Inf or Nan).
Reimplemented from Ipopt::Matrix.
virtual void Ipopt::MultiVectorMatrix::ComputeRowAMaxImpl | ( | Vector & | rows_norms, | |
bool | init | |||
) | const [protected, virtual] |
Method for determining if all stored numbers are valid (i.e., no Inf or Nan).
Implements Ipopt::Matrix.
virtual void Ipopt::MultiVectorMatrix::ComputeColAMaxImpl | ( | Vector & | cols_norms, | |
bool | init | |||
) | const [protected, virtual] |
Method for determining if all stored numbers are valid (i.e., no Inf or Nan).
Implements Ipopt::Matrix.
virtual void Ipopt::MultiVectorMatrix::PrintImpl | ( | const Journalist & | jnlst, | |
EJournalLevel | level, | |||
EJournalCategory | category, | |||
const std::string & | name, | |||
Index | indent, | |||
const std::string & | prefix | |||
) | const [protected, virtual] |
Method for determining if all stored numbers are valid (i.e., no Inf or Nan).
Implements Ipopt::Matrix.
void Ipopt::MultiVectorMatrix::operator= | ( | const MultiVectorMatrix & | ) | [private] |
Overloaded Equals Operator.
Method for accessing the internal Vectors internally.
Definition at line 161 of file IpMultiVectorMatrix.hpp.
Method for accessing the internal Vectors internally.
Definition at line 173 of file IpMultiVectorMatrix.hpp.
const MultiVectorMatrixSpace* Ipopt::MultiVectorMatrix::owner_space_ [private] |
Reimplemented from Ipopt::Matrix.
Definition at line 151 of file IpMultiVectorMatrix.hpp.
std::vector<SmartPtr<const Vector> > Ipopt::MultiVectorMatrix::const_vecs_ [private] |
space for storing the const Vector's
Definition at line 154 of file IpMultiVectorMatrix.hpp.
std::vector<SmartPtr<Vector> > Ipopt::MultiVectorMatrix::non_const_vecs_ [private] |
space for storing the non-const Vector's
Definition at line 157 of file IpMultiVectorMatrix.hpp.