在一些简短的测试中，我发现两者的结合在我所谓的合理密集数学中会更好:

类型:<双[]>列表

时间:00:00:05.1861300

类型：列表<列表<double>>

时间:00:00:05.7941351

类型:double[行*列]

时间:00:00:06.0547118

运行代码:

int rows = 10000;
int columns = 10000;

IMatrix Matrix = new IMatrix(rows, columns);

Stopwatch stopwatch = new Stopwatch();
stopwatch.Start();


for (int r = 0; r < Matrix.Rows; r++)
    for (int c = 0; c < Matrix.Columns; c++)
        Matrix[r, c] = Math.E;

for (int r = 0; r < Matrix.Rows; r++)
    for (int c = 0; c < Matrix.Columns; c++)
        Matrix[r, c] *= -Math.Log(Math.E);


stopwatch.Stop();
TimeSpan ts = stopwatch.Elapsed;

Console.WriteLine(ts.ToString());

我真希望我们有一些顶尖的硬件加速矩阵类，就像。net团队用system . numbers . vectors类做的那样!

c#可能是最好的ML语言，只要在这方面多做一些工作!

static long[] longs = new long[500000];
static long[] longs2 = {};
static List<long> listLongs = new List<long> { };
static void Main(string[] args)
{
    Console.CursorVisible = false;
    Stopwatch time = new Stopwatch();

    time.Start();
    for (int f = 50000000; f < 50255000; f++)
    {
        listLongs.Add(f);
    }

    //List  Time: 1ms    Count : 255000
    Console.WriteLine("List Time: " + time.ElapsedMilliseconds + " | Count: " + listLongs.Count());

    time.Restart();
    time.Start();
    for (long i = 1; i < 500000; i++)
    {
        longs[i] = i * 200;
    }

    //Array Time: 2ms Length: 500000 (Unrealistic Data)
    Console.WriteLine("Array Time: " + time.ElapsedMilliseconds + " | Length: " + longs.Length);

    time.Restart();
    time.Start();
    for (int i = 50000000; i < 50055000; i++)
    {
        longs2 = longs2.Append(i).ToArray();
    }

    //Array Time: 17950ms Length: 55000
    Console.WriteLine("Array Append Time: " + time.ElapsedMilliseconds + " | Length: " + longs2.Length);

    Console.ReadLine();
}

如果您计划不断地向数组中添加少量数据，那么list更快

这实际上取决于你将如何使用数组。

不要试图通过增加元素数量来增加容量。

性能

List For Add: 1ms
Array For Add: 2397ms

    Stopwatch watch;
        #region --> List For Add <--

        List<int> intList = new List<int>();
        watch = Stopwatch.StartNew();
        for (int rpt = 0; rpt < 60000; rpt++)
        {
            intList.Add(rand.Next());
        }
        watch.Stop();
        Console.WriteLine("List For Add: {0}ms", watch.ElapsedMilliseconds);
        #endregion

        #region --> Array For Add <--

        int[] intArray = new int[0];
        watch = Stopwatch.StartNew();
        int sira = 0;
        for (int rpt = 0; rpt < 60000; rpt++)
        {
            sira += 1;
            Array.Resize(ref intArray, intArray.Length + 1);
            intArray[rpt] = rand.Next();

        }
        watch.Stop();
        Console.WriteLine("Array For Add: {0}ms", watch.ElapsedMilliseconds);

        #endregion

很容易测量…

在少量的紧循环处理代码中，我知道长度是固定的，我使用数组来进行额外的微小优化;如果你使用索引器/ for表单，数组可以稍微快一点——但是IIRC认为这取决于数组中数据的类型。但除非你需要进行微观优化，否则尽量保持简单，使用List<T>等。

当然，这只适用于读取所有数据的情况;对于基于键的查找，字典会更快。

下面是我使用“int”的结果(第二个数字是一个校验和，以验证它们都做了相同的工作):

(修改bug)

List/for: 1971ms (589725196)
Array/for: 1864ms (589725196)
List/foreach: 3054ms (589725196)
Array/foreach: 1860ms (589725196)

基于试验台:

using System;
using System.Collections.Generic;
using System.Diagnostics;
static class Program
{
    static void Main()
    {
        List<int> list = new List<int>(6000000);
        Random rand = new Random(12345);
        for (int i = 0; i < 6000000; i++)
        {
            list.Add(rand.Next(5000));
        }
        int[] arr = list.ToArray();

        int chk = 0;
        Stopwatch watch = Stopwatch.StartNew();
        for (int rpt = 0; rpt < 100; rpt++)
        {
            int len = list.Count;
            for (int i = 0; i < len; i++)
            {
                chk += list[i];
            }
        }
        watch.Stop();
        Console.WriteLine("List/for: {0}ms ({1})", watch.ElapsedMilliseconds, chk);

        chk = 0;
        watch = Stopwatch.StartNew();
        for (int rpt = 0; rpt < 100; rpt++)
        {
            for (int i = 0; i < arr.Length; i++)
            {
                chk += arr[i];
            }
        }
        watch.Stop();
        Console.WriteLine("Array/for: {0}ms ({1})", watch.ElapsedMilliseconds, chk);

        chk = 0;
        watch = Stopwatch.StartNew();
        for (int rpt = 0; rpt < 100; rpt++)
        {
            foreach (int i in list)
            {
                chk += i;
            }
        }
        watch.Stop();
        Console.WriteLine("List/foreach: {0}ms ({1})", watch.ElapsedMilliseconds, chk);

        chk = 0;
        watch = Stopwatch.StartNew();
        for (int rpt = 0; rpt < 100; rpt++)
        {
            foreach (int i in arr)
            {
                chk += i;
            }
        }
        watch.Stop();
        Console.WriteLine("Array/foreach: {0}ms ({1})", watch.ElapsedMilliseconds, chk);

        Console.ReadLine();
    }
}

我担心在其他答案中发布的基准测试仍然会为编译器留下优化，消除或合并循环的空间，所以我写了一个:

使用不可预测的输入(随机) 运行计算结果并将结果打印到控制台 每次重复修改输入数据

结果是，直接数组的性能比访问封装在IList中的数组要好250%:

10亿次数组访问:4000毫秒 10亿次列表访问:10000毫秒 1亿个数组访问:350毫秒 1亿次列表访问:1000毫秒

代码如下:

static void Main(string[] args) {
  const int TestPointCount = 1000000;
  const int RepetitionCount = 1000;

  Stopwatch arrayTimer = new Stopwatch();
  Stopwatch listTimer = new Stopwatch();

  Point2[] points = new Point2[TestPointCount];
  var random = new Random();
  for (int index = 0; index < TestPointCount; ++index) {
    points[index].X = random.NextDouble();
    points[index].Y = random.NextDouble();
  }

  for (int repetition = 0; repetition <= RepetitionCount; ++repetition) {
    if (repetition > 0) { // first repetition is for cache warmup
      arrayTimer.Start();
    }
    doWorkOnArray(points);
    if (repetition > 0) { // first repetition is for cache warmup
      arrayTimer.Stop();
    }

    if (repetition > 0) { // first repetition is for cache warmup
      listTimer.Start();
    }
    doWorkOnList(points);
    if (repetition > 0) { // first repetition is for cache warmup
      listTimer.Stop();
    }
  }

  Console.WriteLine("Ignore this: " + points[0].X + points[0].Y);
  Console.WriteLine(
    string.Format(
      "{0} accesses on array took {1} ms",
      RepetitionCount * TestPointCount, arrayTimer.ElapsedMilliseconds
    )
  );
  Console.WriteLine(
    string.Format(
      "{0} accesses on list took {1} ms",
      RepetitionCount * TestPointCount, listTimer.ElapsedMilliseconds
    )
  );

}

private static void doWorkOnArray(Point2[] points) {
  var random = new Random();

  int pointCount = points.Length;

  Point2 accumulated = Point2.Zero;
  for (int index = 0; index < pointCount; ++index) {
    accumulated.X += points[index].X;
    accumulated.Y += points[index].Y;
  }

  accumulated /= pointCount;

  // make use of the result somewhere so the optimizer can't eliminate the loop
  // also modify the input collection so the optimizer can merge the repetition loop
  points[random.Next(0, pointCount)] = accumulated;
}

private static void doWorkOnList(IList<Point2> points) {
  var random = new Random();

  int pointCount = points.Count;

  Point2 accumulated = Point2.Zero;
  for (int index = 0; index < pointCount; ++index) {
    accumulated.X += points[index].X;
    accumulated.Y += points[index].Y;
  }

  accumulated /= pointCount;

  // make use of the result somewhere so the optimizer can't eliminate the loop
  // also modify the input collection so the optimizer can merge the repetition loop
  points[random.Next(0, pointCount)] = accumulated;
}

由于List<>在内部使用数组，因此基本性能应该是相同的。为什么这个列表可能会稍微慢一些，有两个原因:

要在列表中查找元素，调用list方法，该方法在底层数组中进行查找。所以你需要一个额外的方法调用。另一方面，编译器可能会识别出这一点，并优化“不必要的”调用。 如果编译器知道数组的大小，它可能会做一些特殊的优化，而对于一个未知长度的列表，它就不能这样做。如果列表中只有几个元素，这可能会带来一些性能改进。

要检查它是否对您有任何影响，最好将发布的计时函数调整为您计划使用的大小列表，并查看您的特殊情况的结果如何。