I am trying to doing this right now

Code: Select all

Eigen::ArrayXf exposureEfficiencies;
Eigen::MatrixXf m_systemMatrix.getMatrix();


Code: Select all

exposureEfficiencies = (m_systemMatrix.getMatrix().rowwise() * exposureEfficiencies).colwise().sum();


it just won't compile for me

Assuming my matrix has I rows and J columns...
I want to multiply each ROW Matrix(i) of the matrix by the scalar Array(i) and then take the COL-wise sum resulting in a new Array of size J. (YES, this should resize exposureEfficiencies from size I to size J)

Is the following what you're looking for?

Code: Select all

ArrayXXf M(2,3);
M << 1., 2., 3.,
     4., 5., 6.;
ArrayXf A(3);
A << 10., 20., 30.;
   
cout << (M.rowwise() * A.transpose()).colwise().sum() << endl;


which gives

50 140 270.

GeorgeKnight wrote:Is the following what you're looking for?

Code: Select all

ArrayXXf M(2,3);
M << 1., 2., 3.,
     4., 5., 6.;
ArrayXf A(3);
A << 10., 20., 30.;
   
cout << (M.rowwise() * A.transpose()).colwise().sum() << endl;


which gives

50 140 270.

This is exactly what I need.
Although this still isn't compiling....

Code: Select all

exposureEfficiencies = (m_systemMatrix.getMatrix().array().rowwise() * exposureEfficiencies.transpose()).colwise().sum();

Is it possible that ArrayWrapper is missing an overload?

johnm1019 wrote:Although this still isn't compiling....

Code: Select all

exposureEfficiencies = (m_systemMatrix.getMatrix().array().rowwise() * exposureEfficiencies.transpose()).colwise().sum();

Is it possible that ArrayWrapper is missing an overload?

That should compile, assuming that exposureEfficiencies is of type ArrayXf and m_systemMatrix.getMatrix() returns a MatrixXf. Could you perhaps post a small self-contained example showing the error?

The following compiles fine on my computer:

Code: Select all

MatrixXf m_systemMatrix(2,3);
m_systemMatrix << 1., 2., 3.,
                  4., 5., 6.;
ArrayXf exposureEfficiencies(3);
exposureEfficiencies << 10., 20., 30.;

exposureEfficiencies = (m_systemMatrix.array().rowwise() *                         exposureEfficiencies.transpose()).colwise().sum();
       
cout << exposureEfficiencies << endl;


The difference with your code is the "getMatrix()".