Ipopt::DenseVector Class Reference

Dense Vector Implementation. More...

#include <IpDenseVector.hpp>

Inheritance diagram for Ipopt::DenseVector:
Inheritance graph
[legend]
Collaboration diagram for Ipopt::DenseVector:
Collaboration graph
[legend]

List of all members.

Public Member Functions

Constructors / Destructors



 DenseVector (const DenseVectorSpace *owner_space)
 Default Constructor.
virtual ~DenseVector ()
 Destructor.
Additional public methods not in Vector base class.



SmartPtr< DenseVectorMakeNewDenseVector () const
 Create a new DenseVector from same VectorSpace.
void SetValues (const Number *x)
 Set elements in the vector to the Number array x.
NumberValues ()
 Obtain pointer to the internal Number array with vector elements with the indention to change the vector data (USE WITH CARE!).
const NumberValues () const
 Obtain pointer to the internal Number array with vector elements without the intention to change the vector data (USE WITH CARE!).
const NumberExpandedValues () const
 The same as the const version of Values, but we ensure that we always return a valid array, even if IsHomogeneous returns true.
NumberExpandedValues ()
 This is the same as Values, but we add it here so that ExpandedValues can also be used for the non-const case.
bool IsHomogeneous () const
 Indicates if the vector is homogeneous (i.e., all entries have the value Scalar().
Number Scalar () const
 Scalar value of all entries in a homogeneous vector.
Modifying subranges of the vector.



void CopyToPos (Index Pos, const Vector &x)
 Copy the data in x into the subrange of this vector starting at position Pos in this vector.
void CopyFromPos (Index Pos, Vector &x) const
 Copy the data in this vector's subrange starting at position Pos to Vector x.

Protected Member Functions

Overloaded methods from Vector base class



virtual void CopyImpl (const Vector &x)
 Copy the data of the vector x into this vector (DCOPY).
virtual void ScalImpl (Number alpha)
 Scales the vector by scalar alpha (DSCAL).
virtual void AxpyImpl (Number alpha, const Vector &x)
 Add the multiple alpha of vector x to this vector (DAXPY).
virtual Number DotImpl (const Vector &x) const
 Computes inner product of vector x with this (DDOT).
virtual Number Nrm2Impl () const
 Computes the 2-norm of this vector (DNRM2).
virtual Number AsumImpl () const
 Computes the 1-norm of this vector (DASUM).
virtual Number AmaxImpl () const
 Computes the max-norm of this vector (based on IDAMAX).
virtual void SetImpl (Number value)
 Set each element in the vector to the scalar alpha.
virtual void ElementWiseDivideImpl (const Vector &x)
 Element-wise division $y_i \gets y_i/x_i$.
virtual void ElementWiseMultiplyImpl (const Vector &x)
 Element-wise multiplication $y_i \gets y_i*x_i$.
virtual void ElementWiseMaxImpl (const Vector &x)
 Set entry to max of itself and the corresponding element in x.
virtual void ElementWiseMinImpl (const Vector &x)
 Set entry to min of itself and the corresponding element in x.
virtual void ElementWiseReciprocalImpl ()
 reciprocates the elements of the vector
virtual void ElementWiseAbsImpl ()
 take abs of the elements of the vector
virtual void ElementWiseSqrtImpl ()
 take square-root of the elements of the vector
virtual void ElementWiseSgnImpl ()
 Changes each entry in the vector to its sgn value.
virtual void AddScalarImpl (Number scalar)
 Add scalar to every component of the vector.
virtual Number MaxImpl () const
 Max value in the vector.
virtual Number MinImpl () const
 Min value in the vector.
virtual Number SumImpl () const
 Computes the sum of the lements of vector.
virtual Number SumLogsImpl () const
 Computes the sum of the logs of the elements of vector.
Implemented specialized functions



void AddTwoVectorsImpl (Number a, const Vector &v1, Number b, const Vector &v2, Number c)
 Add two vectors (a * v1 + b * v2).
Number FracToBoundImpl (const Vector &delta, Number tau) const
 Fraction to the boundary parameter.
void AddVectorQuotientImpl (Number a, const Vector &z, const Vector &s, Number c)
 Add the quotient of two vectors, y = a * z/s + c * y.
Output methods



virtual void PrintImpl (const Journalist &jnlst, EJournalLevel level, EJournalCategory category, const std::string &name, Index indent, const std::string &prefix) const
 Print the entire vector.

Private Member Functions

Numbervalues_allocated ()
 Method of getting the internal values array, making sure that memory has been allocated.
void set_values_from_scalar ()
 Auxilliary method for setting explicitly all elements in values_ to the current scalar value.
Default Compiler Generated Methods

(Hidden to avoid implicit creation/calling).

These methods are not implemented and we do not want the compiler to implement them for us, so we declare them private and do not define them. This ensures that they will not be implicitly created/called.



 DenseVector ()
 Default Constructor.
 DenseVector (const DenseVector &)
 Copy Constructor.
void operator= (const DenseVector &)
 Overloaded Equals Operator.

Private Attributes

const DenseVectorSpaceowner_space_
 Copy of the owner_space ptr as a DenseVectorSpace instead of a VectorSpace.
Numbervalues_
 Dense Number array of vector values.
Numberexpanded_values_
 Dense Number array pointer that is used for ExpandedValues.
bool initialized_
 Flag for Initialization.
bool homogeneous_
 Flag indicating whether the vector is currently homogeneous (that is, all elements have the same value).
Number scalar_
 Homogeneous value of all elements if the vector is currently homogenous.

Detailed Description

Dense Vector Implementation.

This is the default Vector class in Ipopt. It stores vectors in contiguous Number arrays, unless the vector has the same value in all entires. In the latter case, we call the vector "homogeneous", and we store only the values that is repeated in all elements. If you want to obtain the values of vector, use the IsHomogeneous() method to find out what status the vector is in, and then use either Values() const or Scalar() const methods to get the values. To set the values of a homogeneous method, use the Set method. To set the values of a non-homogeneous vector, use the SetValues method, or use the non-const Values method to get an array that you can overwrite. In the latter case, storage is ensured.

Definition at line 34 of file IpDenseVector.hpp.


Constructor & Destructor Documentation

Ipopt::DenseVector::DenseVector ( const DenseVectorSpace owner_space  ) 

Default Constructor.

virtual Ipopt::DenseVector::~DenseVector (  )  [virtual]

Destructor.

Ipopt::DenseVector::DenseVector (  )  [private]

Default Constructor.

Ipopt::DenseVector::DenseVector ( const DenseVector  )  [private]

Copy Constructor.


Member Function Documentation

SmartPtr< DenseVector > Ipopt::DenseVector::MakeNewDenseVector (  )  const [inline]

Create a new DenseVector from same VectorSpace.

Definition at line 353 of file IpDenseVector.hpp.

void Ipopt::DenseVector::SetValues ( const Number x  ) 

Set elements in the vector to the Number array x.

Number * Ipopt::DenseVector::Values (  )  [inline]

Obtain pointer to the internal Number array with vector elements with the indention to change the vector data (USE WITH CARE!).

This does not produce a copy, and lifetime is not guaranteed!.

Definition at line 304 of file IpDenseVector.hpp.

const Number * Ipopt::DenseVector::Values (  )  const [inline]

Obtain pointer to the internal Number array with vector elements without the intention to change the vector data (USE WITH CARE!).

This does not produce a copy, and lifetime is not guaranteed! IMPORTANT: If this method is currently homogeneous (i.e. IsHomogeneous returns true), then you cannot call this method. Instead, you need to use the Scalar() method.

Definition at line 321 of file IpDenseVector.hpp.

const Number* Ipopt::DenseVector::ExpandedValues (  )  const

The same as the const version of Values, but we ensure that we always return a valid array, even if IsHomogeneous returns true.

Number* Ipopt::DenseVector::ExpandedValues (  )  [inline]

This is the same as Values, but we add it here so that ExpandedValues can also be used for the non-const case.

Definition at line 81 of file IpDenseVector.hpp.

bool Ipopt::DenseVector::IsHomogeneous (  )  const [inline]

Indicates if the vector is homogeneous (i.e., all entries have the value Scalar().

Definition at line 88 of file IpDenseVector.hpp.

Number Ipopt::DenseVector::Scalar (  )  const [inline]

Scalar value of all entries in a homogeneous vector.

Definition at line 94 of file IpDenseVector.hpp.

void Ipopt::DenseVector::CopyToPos ( Index  Pos,
const Vector x 
)

Copy the data in x into the subrange of this vector starting at position Pos in this vector.

Position count starts at 0.

void Ipopt::DenseVector::CopyFromPos ( Index  Pos,
Vector x 
) const

Copy the data in this vector's subrange starting at position Pos to Vector x.

Position count starts at 0.

virtual void Ipopt::DenseVector::CopyImpl ( const Vector x  )  [protected, virtual]

Copy the data of the vector x into this vector (DCOPY).

virtual void Ipopt::DenseVector::ScalImpl ( Number  alpha  )  [protected, virtual]

Scales the vector by scalar alpha (DSCAL).

Implements Ipopt::Vector.

virtual void Ipopt::DenseVector::AxpyImpl ( Number  alpha,
const Vector x 
) [protected, virtual]

Add the multiple alpha of vector x to this vector (DAXPY).

virtual Number Ipopt::DenseVector::DotImpl ( const Vector x  )  const [protected, virtual]

Computes inner product of vector x with this (DDOT).

virtual Number Ipopt::DenseVector::Nrm2Impl (  )  const [protected, virtual]

Computes the 2-norm of this vector (DNRM2).

Implements Ipopt::Vector.

virtual Number Ipopt::DenseVector::AsumImpl (  )  const [protected, virtual]

Computes the 1-norm of this vector (DASUM).

Implements Ipopt::Vector.

virtual Number Ipopt::DenseVector::AmaxImpl (  )  const [protected, virtual]

Computes the max-norm of this vector (based on IDAMAX).

Implements Ipopt::Vector.

virtual void Ipopt::DenseVector::SetImpl ( Number  value  )  [protected, virtual]

Set each element in the vector to the scalar alpha.

Implements Ipopt::Vector.

virtual void Ipopt::DenseVector::ElementWiseDivideImpl ( const Vector x  )  [protected, virtual]

Element-wise division $y_i \gets y_i/x_i$.

virtual void Ipopt::DenseVector::ElementWiseMultiplyImpl ( const Vector x  )  [protected, virtual]

Element-wise multiplication $y_i \gets y_i*x_i$.

virtual void Ipopt::DenseVector::ElementWiseMaxImpl ( const Vector x  )  [protected, virtual]

Set entry to max of itself and the corresponding element in x.

virtual void Ipopt::DenseVector::ElementWiseMinImpl ( const Vector x  )  [protected, virtual]

Set entry to min of itself and the corresponding element in x.

virtual void Ipopt::DenseVector::ElementWiseReciprocalImpl (  )  [protected, virtual]

reciprocates the elements of the vector

Implements Ipopt::Vector.

virtual void Ipopt::DenseVector::ElementWiseAbsImpl (  )  [protected, virtual]

take abs of the elements of the vector

Implements Ipopt::Vector.

virtual void Ipopt::DenseVector::ElementWiseSqrtImpl (  )  [protected, virtual]

take square-root of the elements of the vector

Implements Ipopt::Vector.

virtual void Ipopt::DenseVector::ElementWiseSgnImpl (  )  [protected, virtual]

Changes each entry in the vector to its sgn value.

Implements Ipopt::Vector.

virtual void Ipopt::DenseVector::AddScalarImpl ( Number  scalar  )  [protected, virtual]

Add scalar to every component of the vector.

Implements Ipopt::Vector.

virtual Number Ipopt::DenseVector::MaxImpl (  )  const [protected, virtual]

Max value in the vector.

Implements Ipopt::Vector.

virtual Number Ipopt::DenseVector::MinImpl (  )  const [protected, virtual]

Min value in the vector.

Implements Ipopt::Vector.

virtual Number Ipopt::DenseVector::SumImpl (  )  const [protected, virtual]

Computes the sum of the lements of vector.

Implements Ipopt::Vector.

virtual Number Ipopt::DenseVector::SumLogsImpl (  )  const [protected, virtual]

Computes the sum of the logs of the elements of vector.

Implements Ipopt::Vector.

void Ipopt::DenseVector::AddTwoVectorsImpl ( Number  a,
const Vector v1,
Number  b,
const Vector v2,
Number  c 
) [protected]

Add two vectors (a * v1 + b * v2).

Result is stored in this vector.

Number Ipopt::DenseVector::FracToBoundImpl ( const Vector delta,
Number  tau 
) const [protected]

Fraction to the boundary parameter.

void Ipopt::DenseVector::AddVectorQuotientImpl ( Number  a,
const Vector z,
const Vector s,
Number  c 
) [protected]

Add the quotient of two vectors, y = a * z/s + c * y.

virtual void Ipopt::DenseVector::PrintImpl ( const Journalist jnlst,
EJournalLevel  level,
EJournalCategory  category,
const std::string &  name,
Index  indent,
const std::string &  prefix 
) const [protected, virtual]

Print the entire vector.

Implements Ipopt::Vector.

void Ipopt::DenseVector::operator= ( const DenseVector  )  [private]

Overloaded Equals Operator.

Reimplemented from Ipopt::Vector.

Number * Ipopt::DenseVector::values_allocated (  )  [inline, private]

Method of getting the internal values array, making sure that memory has been allocated.

Definition at line 327 of file IpDenseVector.hpp.

void Ipopt::DenseVector::set_values_from_scalar (  )  [private]

Auxilliary method for setting explicitly all elements in values_ to the current scalar value.


Member Data Documentation

Copy of the owner_space ptr as a DenseVectorSpace instead of a VectorSpace.

Reimplemented from Ipopt::Vector.

Definition at line 225 of file IpDenseVector.hpp.

Dense Number array of vector values.

Definition at line 228 of file IpDenseVector.hpp.

Dense Number array pointer that is used for ExpandedValues.

Definition at line 231 of file IpDenseVector.hpp.

Flag for Initialization.

This flag is false, if the data has not yet been initialized.

Definition at line 239 of file IpDenseVector.hpp.

Flag indicating whether the vector is currently homogeneous (that is, all elements have the same value).

This flag is used to determine whether the elements of the vector are stored in values_ or in scalar_

Definition at line 245 of file IpDenseVector.hpp.

Homogeneous value of all elements if the vector is currently homogenous.

Definition at line 249 of file IpDenseVector.hpp.


The documentation for this class was generated from the following file:

Generated on 15 Mar 2015 for Coin-All by  doxygen 1.6.1