我正在尝试在Eigen中实现Matlab的m(:,v> 0)和m(:,[1,3])。到目前为止,我有以下功能:
MatrixXd select_cols(const MatrixXd src, const Matrix<bool, 1, Dynamic> cond)
{
MatrixXd dst(src.rows(), cond.count());
unsigned int i = 0;
for (unsigned int j = 0; j < cond.cols(); j++) {
if (cond(0,j)) {
dst.col(i++) = src.col(j);
}
}
return dst;
}
MatrixXd select_cols(const MatrixXd src, const Matrix<unsigned int, 1, Dynamic> idx)
{
MatrixXd dst(src.rows(), idx.size());
for (unsigned int i = 0; i < idx.size(); ++i) {
dst.col(i) = src.col(idx[i]);
}
return dst;
}
但是,我认为以下代码段可行,但它不会:
Eigen::MatrixXd m1(3,3);
Eigen::Matrix<unsigned int, 1, Eigen::Dynamic> v(3);
m1 << 1,2,3,
4,5,6,
7,8,9;
v << 2,0,2;
Eigen::MatrixXd m2 = Eigen::select_cols(m1, v.array() > 1);
有错误:
utest.cpp:88:62: error: call of overloaded ‘select_cols(Eigen::MatrixXd&, const Eigen::CwiseUnaryOp<std::binder2nd<std::greater<unsigned int> >, const Eigen::ArrayWrapper<Eigen::Matrix<unsigned int, 1, -0x00000000000000001> > >)’ is ambiguous
utest.cpp:88:62: note: candidates are:
common.h:29:14: note: Eigen::MatrixXd Eigen::select_cols(Eigen::MatrixXd, Eigen::Matrix<unsigned int, 1, -0x00000000000000001>)
common.h:32:14: note: Eigen::MatrixXd Eigen::select_cols(Eigen::MatrixXd, Eigen::Matrix<bool, 1, -0x00000000000000001>)
答案 0 :(得分:0)
要在Eigen中复制Matlab的 m(:,[1,3]),您只需使用矩阵类中的 .block()方法:
MatrixXd m(3,4);
m << 1,2,3,4,5,6,7,8,9,10,11,12;
...
MatrixXd sub_m = m.block(0,0, m.rows(),3);
...
std::cout << m << std::endl << sub_m << std::endl;
输出将是:
1 2 3 4
5 6 7 8
9 10 11 12
1 2 3
5 6 7
9 10 11
问候。