指向本征矩阵的指针数组

指向本征矩阵的指针数组

本文介绍了指向本征矩阵的指针数组的处理方法,对大家解决问题具有一定的参考价值,需要的朋友们下面随着小编来一起学习吧!

问题描述

我在代码上使用了Eigen的MatrixXd矩阵,在某个时候我需要3D矩阵.由于Eigen没有三维矩阵类型,因为它仅针对线性代数进行了优化,所以我创建了MatrixXd类型的指针数组:

I am using MatrixXd matrices from Eigen on my code, and at a certain point I need a 3D one. Since Eigen does not have tridimensional matrix types, as it is optimized just for linear algebra, instead I am creating a pointer array of the MatrixXd type:

Eigen::MatrixXd *CVM =new Eigen::MatrixXd[100];
for (int i = 0; i < 100; i++){
   CVM[i]= Eigen::MatrixXd::Zero(5,5);
}

但是,稍后我需要访问该数组上的值,为此,我正在做类似的事情:

However, later on I need to acess the values on this array, and for that I am doing something like:

for (int k = 0; k < 100; k++){
   Eigen::MatrixXd* b=&CVM[k];

   for (int i = 0; i < 5; i++){
      for (int j = 0; j < 5; j++){
         b->coeff(i,j)=47;
      }
   }
}

由于b是指针而不是MatrixXd本身,所以b(i,j)显然不起作用,所以我改用coeff()方法,但是,出现以下错误:

As b is a pointer and not the MatrixXd itself, b(i,j) obviously wouldn't work, so instead I am using the coeff() method, however, I get the following error:

error: assignment of read-only location ‘b->Eigen::Matrix<double, -1, -1>::<anonymous>.Eigen::PlainObjectBase<Derived>::coeff<Eigen::Matrix<double, -1, -1> >(((Eigen::PlainObjectBase<Eigen::Matrix<double, -1, -1> >::Index)i), ((Eigen::PlainObjectBase<Eigen::Matrix<double, -1, -1> >::Index)j))’

添加了输出

cout << b << endl;
cout << CVM[47] << endl;


0x1c34b00
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0

推荐答案

要么使用b->coeffRef(i,j)获取读/写引用,要么取消引用b:(*b)(i,j),或者为b使用引用:

Either use b->coeffRef(i,j) to get a read/write reference, or dereference b: (*b)(i,j), or use a reference for b:

MatrixXd& b = CVM[k];
b(i,j) = 47;

这篇关于指向本征矩阵的指针数组的文章就介绍到这了,希望我们推荐的答案对大家有所帮助,也希望大家多多支持!

08-19 23:43