template<typename XprType, int BlockRows, int BlockCols, bool InnerPanel>
class Eigen::Block< XprType, BlockRows, BlockCols, InnerPanel >

Expression of a fixed-size or dynamic-size block.

Template Parameters

XprType	the type of the expression in which we are taking a block
BlockRows	the number of rows of the block we are taking at compile time (optional)
BlockCols	the number of columns of the block we are taking at compile time (optional)
InnerPanel	is true, if the block maps to a set of rows of a row major matrix or to set of columns of a column major matrix (optional). The parameter allows to determine at compile time whether aligned access is possible on the block expression.

This class represents an expression of either a fixed-size or dynamic-size block. It is the return type of DenseBase::block(Index,Index,Index,Index) and DenseBase::block<int,int>(Index,Index) and most of the time this is the only way it is used.

However, if you want to directly maniputate block expressions, for instance if you want to write a function returning such an expression, you will need to use this class.

Here is an example illustrating the dynamic case:

#include <Eigen/Core>
#include <iostream>
using namespace Eigen;
using namespace std;
 
template<typename Derived>
Eigen::Block<Derived>
topLeftCorner(MatrixBase<Derived>& m, int rows, int cols)
{
  return Eigen::Block<Derived>(m.derived(), 0, 0, rows, cols);
}
 
template<typename Derived>
const Eigen::Block<const Derived>
topLeftCorner(const MatrixBase<Derived>& m, int rows, int cols)
{
  return Eigen::Block<const Derived>(m.derived(), 0, 0, rows, cols);
}
 
int main(int, char**)
{
  Matrix4d m = Matrix4d::Identity();
  cout << topLeftCorner(4*m, 2, 3) << endl; // calls the const version
  topLeftCorner(m, 2, 3) *= 5;              // calls the non-const version
  cout << "Now the matrix m is:" << endl << m << endl;
  return 0;
}

Output:

4 0 0
0 4 0
Now the matrix m is:
5 0 0 0
0 5 0 0
0 0 1 0
0 0 0 1

Note

Even though this expression has dynamic size, in the case where XprType has fixed size, this expression inherits a fixed maximal size which means that evaluating it does not cause a dynamic memory allocation.

Here is an example illustrating the fixed-size case:

#include <Eigen/Core>
#include <iostream>
using namespace Eigen;
using namespace std;
 
template<typename Derived>
Eigen::Block<Derived, 2, 2>
topLeft2x2Corner(MatrixBase<Derived>& m)
{
  return Eigen::Block<Derived, 2, 2>(m.derived(), 0, 0);
}
 
template<typename Derived>
const Eigen::Block<const Derived, 2, 2>
topLeft2x2Corner(const MatrixBase<Derived>& m)
{
  return Eigen::Block<const Derived, 2, 2>(m.derived(), 0, 0);
}
 
int main(int, char**)
{
  Matrix3d m = Matrix3d::Identity();
  cout << topLeft2x2Corner(4*m) << endl; // calls the const version
  topLeft2x2Corner(m) *= 2;              // calls the non-const version
  cout << "Now the matrix m is:" << endl << m << endl;
  return 0;
}

Output:

4 0
0 4
Now the matrix m is:
2 0 0
0 2 0
0 0 1

See also

DenseBase::block(Index,Index,Index,Index), DenseBase::block(Index,Index), class VectorBlock