嗯，这可能是非常基本的问题（我觉得杀我）......但

什么是交错数组和多维数组之间的差异。 有一个在另一个的好处？

和为什么会在Visual Studio不允许我做了

  MyClass的[] [] ABC =新MyClass的[10] [20];
 

（我们用来做，在C ++，但在C＃它强调[20]用红色wriggly行..说无效排名符）​​

但高兴与

  MyClass的[，] ABC =新MyClass的[10,20]。
 

最后，我怎么能初始化这一行（就像我们做简单的数组与 {新的XXX ...} {新的XXX ....} ）

  MyClass的[] [] [，] itemscollection;
 

解决方案

一个交错数组是一个数组，数组的-，所以一个 INT [] [] 是一个数组 INT [] ，其中每一个可以是不同长度的，并占据其自己的块在存储器中。多维数组（ INT [，] ）是一个单独的内存块（本质上是一个矩阵）。

您不能创建一个 MyClass的[10] [20] ，因为每个子阵列必须单独初始化，因为它们是独立的对象：

  MyClass的[] [] ABC =新MyClass的[10] [];

的for（int i = 0; I＆LT; abc.Length;我++）{
    ABC [我] =新MyClass的[20]。
}
 

在 MyClass的[10,20] 是确定的，因为它正在初始化一个对象为10行和20列的矩阵

A MyClass的[] [] [，] 可initailized像这样（不编译尽管测试）：

  MyClass的[] [] [，] ABC =新MyClass的[10] [，] [，]

的for（int i = 0; I＆LT; abc.Length;我++）{
    ABC [我] =新MyClass的[20,30] [，]

    为（诠释J = 0; J＆其中; ABC [I] .GetLength（0）; J ++）{
        为（中间体K = 0; K＆其中; ABC [I] .GetLength（1）; k ++）{
            ABC [I] [J，K] =新MyClass的[40,50]。
        }
    }
}
 

记住，CLR的是大量的一维数组访问进行了优化，所以采用交错数组可能会比同样大小的多维数组更快。

Well this may be very basic question (I feel like killing myself)... But

What is the difference between jagged array and Multidimensional array. Is there a benefit of one on another?

And why would the Visual Studio not allow me to do a

MyClass[][] abc = new MyClass[10][20];

(We used to do that in C++, but in C# it underlines [20] with red wriggly line.. Says invalid rank specifier)

but is happy with

MyClass[,] abc = new MyClass[10,20];

Finally how can I initialize this in a single line (like we do in simple array with {new xxx...}{new xxx....})

MyClass[][,][,] itemscollection;

A jagged array is an array-of-arrays, so an int[][] is an array of int[], each of which can be of different lengths and occupy their own block in memory. A multidimensional array (int[,]) is a single block of memory (essentially a matrix).

You can't create a MyClass[10][20] because each sub-array has to be initialized separately, as they are separate objects:

MyClass[][] abc = new MyClass[10][];

for (int i=0; i<abc.Length; i++) {
    abc[i] = new MyClass[20];
}

a MyClass[10,20] is ok because it is initializing a single object as a matrix with 10 rows and 20 columns

A MyClass[][,][,] can be initailized like so (not compile tested though):

MyClass[][,][,] abc = new MyClass[10][,][,];

for (int i=0; i<abc.Length; i++) {
    abc[i] = new MyClass[20,30][,];

    for (int j=0; j<abc[i].GetLength(0); j++) {
        for (int k=0; k<abc[i].GetLength(1); k++) {
            abc[i][j,k] = new MyClass[40,50];
        }
    }
}

Bear in mind that the CLR is heavily optimized for single-dimension array access, so using a jagged array will likely be faster than a multidimensional array of the same size.