表达模板和矩阵乘法 [英] Expression templates and matrix multiplication

问题描述

您好，



我正在尝试理解使用表达式模板技术（如Eigen for exmaple）的数学库如何处理矩阵乘法。



我尝试了什么：



在很多例子中，我发现总是只有元素-wise操作实现：



A = B * C + D



编译为单循环



for（i = 0; i< size; i ++）

{

A [i] = B [i] * C [i] + D [i];

}



但如何管理矩阵乘法（需要嵌套循环）？



谢谢

Hello,

I'm trying to understand how mathematical libraries that use Expression Templates techniques ( like Eigen for exmaple ) handle Matrix Multiplication.

What I have tried:

In many examples I found there is always just the element-wise operation implemented:

A = B*C + D

is compiled as a single loop

for ( i = 0 ; i < size; i++ )
{
A[i] = B[i]*C[i] + D[i];
}

but how to manage matrix multiplication ( that needs nested loops )?

Thank you

推荐答案

矩阵乘法c - Google搜索 [ ^ ]

我终于找到一个链接我从未注意到我的问题的答案





Eigen：懒惰评估和别名 [ ^ ]

I finally found a link i never noticed that answer to my question


Eigen: Lazy Evaluation and Aliasing[^]

这篇关于表达模板和矩阵乘法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IT屋！