c++中用vector创建多维数组的初始化方法

最近调试一个程序，在使用vector声明一个二维数组时出现错误。错误的方法如下所示：

std::vector<std::vector<double> > sphereGrid;
int gridLA = angleSpanLA / angelAccuracy;
int gridLO = angleSpanLO / angelAccuracy;
sphereGrid = std::vector<std::vector<double> >( gridLA , gridLO );

会出现如下报错：

/home/zn/VanishingPointDetection/src/VPDetection.cpp: In member function ‘void VPDetection::getSphereGrids(std::vector<std::vector<double> >&)’:
/home/zn/VanishingPointDetection/src/VPDetection.cpp:156:66: error: no matching function for call to ‘std::vector<std::vector<double> >::vector(int&, int&)’
  sphereGrid = std::vector<std::vector<double> >( gridLA , gridLO );

这就是因为二维数组的初始化出现了错误，一般的话要通过下面这种方式初始化

定义空二维vector，再赋值
vector<vector <int> > ivec(m ,vector<int>(n)); //m*n的二维vector，注意两个 "> "之间要有空格！

所以我们要把程序改为

std::vector<std::vector<double> > sphereGrid;
int gridLA = angleSpanLA / angelAccuracy;
int gridLO = angleSpanLO / angelAccuracy;
sphereGrid = std::vector<std::vector<double> >( gridLA , std::vector<double>(gridLO)  );

就可以解决错误，通过这次改错更加认识到了c++之vector的用法。

参考：https://blog.csdn.net/ldkcumt/article/details/51396980

https://blog.csdn.net/oNever_say_love/article/details/50763238