盘旋成螺旋状

一个朋友需要一种算法，该算法可以让他遍历NxM矩阵的元素（N和M为奇数）。我想出了一个解决方案，但我想看看我的SO同事是否可以想出更好的解决方案。

我将我的解决方案发布为该问题的答案。

示例输出：

对于3x3矩阵，输出应为：

（0，0）（1，0）（1，1）（0，1）（-1，1）（-1，0）（-1，-1）（0，-1）（1，-1 ）

此外，该算法应支持非平方矩阵，因此对于5x3矩阵，输出应为：

（0，0）（1，0）（1，1）（0，1）（-1，1）（-1，0）（-1，-1）（0，-1）（1，-1 ）（2，-1）（2，0）（2，1）（-2，1）（-2，0）（-2，-1）

这是我的解决方案（在Python中）：

def spiral(X, Y):
    x = y = 0
    dx = 0
    dy = -1
    for i in range(max(X, Y)**2):
        if (-X/2 < x <= X/2) and (-Y/2 < y <= Y/2):
            print (x, y)
            # DO STUFF...
        if x == y or (x < 0 and x == -y) or (x > 0 and x == 1-y):
            dx, dy = -dy, dx
        x, y = x+dx, y+dy

C ++有人吗？从python快速翻译，发布以保持完整性

void Spiral( int X, int Y){
    int x,y,dx,dy;
    x = y = dx =0;
    dy = -1;
    int t = std::max(X,Y);
    int maxI = t*t;
    for(int i =0; i < maxI; i++){
        if ((-X/2 <= x) && (x <= X/2) && (-Y/2 <= y) && (y <= Y/2)){
            // DO STUFF...
        }
        if( (x == y) || ((x < 0) && (x == -y)) || ((x > 0) && (x == 1-y))){
            t = dx;
            dx = -dy;
            dy = t;
        }
        x += dx;
        y += dy;
    }
}

let x = 0
let y = 0
let d = 1
let m = 1

while true
  while 2 * x * d < m
    print(x, y)
    x = x + d
  while 2 * y * d < m
    print(x, y)
    y = y + d
  d = -1 * d
  m = m + 1

对于以各种编程语言编写的该问题，已经提出了许多建议的解决方案，但是它们似乎都源自相同的复杂方法。我将考虑计算螺旋的更一般的问题，可以使用归纳法将其简洁地表达出来。

基本情况：从（0，0）开始，向前移动1个正方形，向左转，向前移动1个正方形，向左转。归纳步骤：向前移动n + 1个正方形，向左转，向前移动n + 1个正方形，向左转。

表达此问题的数学优雅性强烈建议应该有一个简单的算法来计算解决方案。牢记抽象，我选择不以特定的编程语言实现算法，而是以伪代码实现。

首先，我将考虑一种算法，该算法使用4对while循环来计算螺旋的2次迭代。每对的结构相似，但其自身却各不相同。乍一看似乎很疯狂（某些循环仅执行一次），但我将逐步进行转换，直到我们得到4对相同的循环，因此可以用放置在另一个循环内的一对替换。这将为我们提供不使用任何条件即可计算n次迭代的一般解决方案。

let x = 0
let y = 0

//RIGHT, UP
while x < 1
  print(x, y)
  x = x + 1
while y < 1
  print(x, y)
  y = y + 1

//LEFT, LEFT, DOWN, DOWN
while x > -1
  print(x, y)
  x = x - 1
while y > -1
  print(x, y)
  y = y - 1

//RIGHT, RIGHT, RIGHT, UP, UP, UP
while x < 2
  print(x, y)
  x = x + 1
while y < 2
  print(x, y)
  y = y + 1

//LEFT, LEFT, LEFT, LEFT, DOWN, DOWN, DOWN, DOWN
while x > -2
  print(x, y)
  x = x - 1
while y > -2
  print(x, y)
  y = y - 1

我们将进行的第一个转换是引入一个新的变量d，该变量的方向为+1或-1。方向在每对循环之后切换。由于我们知道d在所有点上的值，因此我们可以将每个不等式的每一边乘以它，相应地调整不等式的方向，并将d与常数的任何乘积简化为另一个常数。这给我们留下了以下内容。

let x = 0
let y = 0
let d = 1

//RIGHT, UP
while x * d < 1
  print(x, y)
  x = x + d
while y * d < 1
  print(x, y)
  y = y + d
d = -1 * d

//LEFT, LEFT, DOWN, DOWN
while x * d < 1
  print(x, y)
  x = x + d
while y * d < 1
  print(x, y)
  y = y + d
d = -1 * d

//RIGHT, RIGHT, RIGHT, UP, UP, UP
while x * d < 2
  print(x, y)
  x = x + d
while y * d < 2
  print(x, y)
  y = y + d
d = -1 * d

//LEFT, LEFT, LEFT, LEFT, DOWN, DOWN, DOWN, DOWN
while x * d < 2
  print(x, y)
  x = x + d
while y * d < 2
  print(x, y)
  y = y + d

现在我们注意到x * d和RHS都是整数，因此我们可以从RHS中减去0到1之间的任何实数值，而不会影响不等式的结果。我们选择从每对while循环对的不等式中减去0.5，以建立更多的模式。

let x = 0
let y = 0
let d = 1

//RIGHT, UP
while x * d < 0.5
  print(x, y)
  x = x + d
while y * d < 0.5
  print(x, y)
  y = y + d
d = -1 * d

//LEFT, LEFT, DOWN, DOWN
while x * d < 1
  print(x, y)
  x = x + d
while y * d < 1
  print(x, y)
  y = y + d
d = -1 * d

//RIGHT, RIGHT, RIGHT, UP, UP, UP
while x * d < 1.5
  print(x, y)
  x = x + d
while y * d < 1.5
  print(x, y)
  y = y + d
d = -1 * d

//LEFT, LEFT, LEFT, LEFT, DOWN, DOWN, DOWN, DOWN
while x * d < 2
  print(x, y)
  x = x + d
while y * d < 2
  print(x, y)
  y = y + d

现在，我们可以为每个while循环对采取的步数引入另一个变量m。

let x = 0
let y = 0
let d = 1
let m = 0.5

//RIGHT, UP
while x * d < m
  print(x, y)
  x = x + d
while y * d < m
  print(x, y)
  y = y + d
d = -1 * d
m = m + 0.5

//LEFT, LEFT, DOWN, DOWN
while x * d < m
  print(x, y)
  x = x + d
while y * d < m
  print(x, y)
  y = y + d
d = -1 * d
m = m + 0.5

//RIGHT, RIGHT, RIGHT, UP, UP, UP
while x * d < m
  print(x, y)
  x = x + d
while y * d < m
  print(x, y)
  y = y + d
d = -1 * d
m = m + 0.5

//LEFT, LEFT, LEFT, LEFT, DOWN, DOWN, DOWN, DOWN
while x * d < m
  print(x, y)
  x = x + d
while y * d < m
  print(x, y)
  y = y + d

最后，我们看到，每对while循环的结构相同，可以简化为放置在另一个循环内部的单个循环。另外，为避免使用实数值，我将m的初始值乘以；m值增加；每个不等式的两边都加2。

这导致该答案开头显示的解决方案。

这是一个O（1）解决方案，用于找到平方螺旋形的位置：小提琴

function spiral(n) {
    // given n an index in the squared spiral
    // p the sum of point in inner square
    // a the position on the current square
    // n = p + a

    var r = Math.floor((Math.sqrt(n + 1) - 1) / 2) + 1;

    // compute radius : inverse arithmetic sum of 8+16+24+...=
    var p = (8 * r * (r - 1)) / 2;
    // compute total point on radius -1 : arithmetic sum of 8+16+24+...

    var en = r * 2;
    // points by face

    var a = (1 + n - p) % (r * 8);
    // compute de position and shift it so the first is (-r,-r) but (-r+1,-r)
    // so square can connect

    var pos = [0, 0, r];
    switch (Math.floor(a / (r * 2))) {
        // find the face : 0 top, 1 right, 2, bottom, 3 left
        case 0:
            {
                pos[0] = a - r;
                pos[1] = -r;
            }
            break;
        case 1:
            {
                pos[0] = r;
                pos[1] = (a % en) - r;

            }
            break;
        case 2:
            {
                pos[0] = r - (a % en);
                pos[1] = r;
            }
            break;
        case 3:
            {
                pos[0] = -r;
                pos[1] = r - (a % en);
            }
            break;
    }
    console.log("n : ", n, " r : ", r, " p : ", p, " a : ", a, "  -->  ", pos);
    return pos;
}

我喜欢python的生成器。

def spiral(N, M):
    x,y = 0,0   
    dx, dy = 0, -1

    for dumb in xrange(N*M):
        if abs(x) == abs(y) and [dx,dy] != [1,0] or x>0 and y == 1-x:  
            dx, dy = -dy, dx            # corner, change direction

        if abs(x)>N/2 or abs(y)>M/2:    # non-square
            dx, dy = -dy, dx            # change direction
            x, y = -y+dx, x+dy          # jump

        yield x, y
        x, y = x+dx, y+dy

测试：

print 'Spiral 3x3:'
for a,b in spiral(3,3):
    print (a,b),

print '\n\nSpiral 5x3:'
for a,b in spiral(5,3):
    print (a,b),

你得到：

Spiral 3x3:
(0, 0) (1, 0) (1, 1) (0, 1) (-1, 1) (-1, 0) (-1, -1) (0, -1) (1, -1) 

Spiral 5x3:
(0, 0) (1, 0) (1, 1) (0, 1) (-1, 1) (-1, 0) (-1, -1) (0, -1) (1, -1) (2, -1) (2, 0) (2, 1) (-2, 1) (-2, 0) (-2, -1)

基于C ++变体的Java螺旋式“代码高尔夫”尝试。

public static void Spiral(int X, int Y) {
    int x=0, y=0, dx = 0, dy = -1;
    int t = Math.max(X,Y);
    int maxI = t*t;

    for (int i=0; i < maxI; i++){
        if ((-X/2 <= x) && (x <= X/2) && (-Y/2 <= y) && (y <= Y/2)) {
            System.out.println(x+","+y);
            //DO STUFF
        }

        if( (x == y) || ((x < 0) && (x == -y)) || ((x > 0) && (x == 1-y))) {
            t=dx; dx=-dy; dy=t;
        }   
        x+=dx; y+=dy;
    }
}

这是一个C ++解决方案，它显示您可以直接且轻松地从前一个坐标计算下一个（x，y）坐标-无需跟踪当前方向，半径或其他任何值：

void spiral(const int M, const int N)
{
    // Generate an Ulam spiral centered at (0, 0).
    int x = 0;
    int y = 0;

    int end = max(N, M) * max(N, M);
    for(int i = 0; i < end; ++i)
    {
        // Translate coordinates and mask them out.
        int xp = x + N / 2;
        int yp = y + M / 2;
        if(xp >= 0 && xp < N && yp >= 0 && yp < M)
            cout << xp << '\t' << yp << '\n';

        // No need to track (dx, dy) as the other examples do:
        if(abs(x) <= abs(y) && (x != y || x >= 0))
            x += ((y >= 0) ? 1 : -1);
        else
            y += ((x >= 0) ? -1 : 1);
    }
}

如果您要做的只是在螺旋中生成前N个点（没有原始问题的屏蔽范围，即对N x M区域的约束），代码将变得非常简单：

void spiral(const int N)
{
    int x = 0;
    int y = 0;
    for(int i = 0; i < N; ++i)
    {
        cout << x << '\t' << y << '\n';
        if(abs(x) <= abs(y) && (x != y || x >= 0))
            x += ((y >= 0) ? 1 : -1);
        else
            y += ((x >= 0) ? -1 : 1);
    }
}

诀窍是您可以比较x和y来确定所处正方形的哪一侧，并告诉您移动的方向。

TDD，使用Java。

SpiralTest.java：

import java.awt.Point;
import java.util.List;

import junit.framework.TestCase;

public class SpiralTest extends TestCase {

    public void test3x3() throws Exception {
        assertEquals("(0, 0) (1, 0) (1, 1) (0, 1) (-1, 1) (-1, 0) (-1, -1) (0, -1) (1, -1)", strung(new Spiral(3, 3).spiral()));
    }

    public void test5x3() throws Exception {
        assertEquals("(0, 0) (1, 0) (1, 1) (0, 1) (-1, 1) (-1, 0) (-1, -1) (0, -1) (1, -1) (2, -1) (2, 0) (2, 1) (-2, 1) (-2, 0) (-2, -1)",
                strung(new Spiral(5, 3).spiral()));
    }

    private String strung(List<Point> points) {
        StringBuffer sb = new StringBuffer();
        for (Point point : points)
            sb.append(strung(point));
        return sb.toString().trim();
    }

    private String strung(Point point) {
        return String.format("(%s, %s) ", point.x, point.y);
    }

}

Spiral.java：

import java.awt.Point;
import java.util.ArrayList;
import java.util.List;

public class Spiral {
    private enum Direction {
    E(1, 0) {Direction next() {return N;}},
    N(0, 1) {Direction next() {return W;}},
    W(-1, 0) {Direction next() {return S;}},
    S(0, -1) {Direction next() {return E;}},;

        private int dx;
        private int dy;

        Point advance(Point point) {
            return new Point(point.x + dx, point.y + dy);
        }

        abstract Direction next();

        Direction(int dx, int dy) {
            this.dx = dx;
            this.dy = dy;
        }
    };
    private final static Point ORIGIN = new Point(0, 0);
    private final int   width;
    private final int   height;
    private Point       point;
    private Direction   direction   = Direction.E;
    private List<Point> list = new ArrayList<Point>();

    public Spiral(int width, int height) {
        this.width = width;
        this.height = height;
    }

    public List<Point> spiral() {
        point = ORIGIN;
        int steps = 1;
        while (list.size() < width * height) {
            advance(steps);
            advance(steps);
            steps++;
        }
        return list;
    }

    private void advance(int n) {
        for (int i = 0; i < n; ++i) {
            if (inBounds(point))
                list.add(point);
            point = direction.advance(point);
        }
        direction = direction.next();
    }

    private boolean inBounds(Point p) {
        return between(-width / 2, width / 2, p.x) && between(-height / 2, height / 2, p.y);
    }

    private static boolean between(int low, int high, int n) {
        return low <= n && n <= high;
    }
}

这是我的解决方案（在Ruby中）

def spiral(xDim, yDim)
   sx = xDim / 2
   sy = yDim / 2

   cx = cy = 0
   direction = distance = 1

   yield(cx,cy)
   while(cx.abs <= sx || cy.abs <= sy)
      distance.times { cx += direction; yield(cx,cy) if(cx.abs <= sx && cy.abs <= sy); } 
      distance.times { cy += direction; yield(cx,cy) if(cx.abs <= sx && cy.abs <= sy); } 
      distance += 1
      direction *= -1
   end
end

spiral(5,3) { |x,y|
   print "(#{x},#{y}),"
}

Haskell，请选择：

spiral x y = (0, 0) : concatMap ring [1 .. max x' y'] where
    ring n | n > x' = left x' n  ++ right x' (-n)
    ring n | n > y' = up   n  y' ++ down (-n) y'
    ring n          = up n n ++ left n n ++ down n n ++ right n n
    up    x y = [(x, n) | n <- [1-y .. y]]; down = (.) reverse . up
    right x y = [(n, y) | n <- [1-x .. x]]; left = (.) reverse . right
    (x', y') = (x `div` 2, y `div` 2)

spiral x y = filter (\(x',y') -> 2*abs x' <= x && 2*abs y' <= y) .
             scanl (\(a,b) (c,d) -> (a+c,b+d)) (0,0) $
             concat [ (:) (1,0) . tail 
                    $ concatMap (replicate n) [(0,1),(-1,0),(0,-1),(1,0)]
                    | n <- [2,4..max x y] ]

这是C语言。

我碰巧选择了错误的变量名。在名称T ==顶部，L ==左侧，B ==底部，R ==右侧。因此，tli是i的左上方，而brj是j的右下方。

#include<stdio.h>

typedef enum {
   TLTOR = 0,
   RTTOB,
   BRTOL,
   LBTOT
} Direction;

int main() {
   int arr[][3] = {{1,2,3},{4,5,6}, {7,8,9}, {10,11,12}};
   int tli = 0, tlj = 0, bri = 3, brj = 2;
   int i;
   Direction d = TLTOR;

   while (tli < bri || tlj < brj) {
     switch (d) {
     case TLTOR:
    for (i = tlj; i <= brj; i++) {
       printf("%d ", arr[tli][i]);
    }
    tli ++;
    d = RTTOB;
    break;
     case RTTOB:
    for (i = tli; i <= bri; i++) {
       printf("%d ", arr[i][brj]);
    }
    brj --;
    d = BRTOL;
    break;
     case BRTOL:
    for (i = brj; i >= tlj; i--) {
       printf("%d ", arr[bri][i]);
    }
    bri --;
        d = LBTOT;
    break;
     case LBTOT:
    for (i = bri; i >= tli; i--) {
       printf("%d ", arr[i][tlj]);
    }
    tlj ++;
        d = TLTOR;
    break;
 }
   }
   if (tli == bri == tlj == brj) {
      printf("%d\n", arr[tli][tlj]);
   }
}

我有一个开源库pixelcan，这是一个python库，它提供功能以多种空间模式扫描网格上的像素。包括的空间模式是圆形，环形，网格，蛇形和随机游走。也有各种变换（例如，剪辑，交换，旋转，平移）。原来的OP问题可以解决如下

for x, y in clip(swap(ringscan(0, 0, 0, 2)), miny=-1, maxy=1):
    print x, y

这产生了点

(0,0) (1,0) (1,1) (0,1) (-1,1) (-1,0) (-1,-1) (0,-1) (1,-1) (2,0) (2,1) (-2,1) (-2,0)
(-2,-1) (2,-1)

库生成器和变换可以链接在一起，以多种顺序和空间模式来更改点。

这是Python 3中的一种解决方案，用于按顺时针和逆时针螺旋方式打印连续的整数。

import math

def sp(n): # spiral clockwise
    a=[[0 for x in range(n)] for y in range(n)]
    last=1
    for k in range(n//2+1):
      for j in range(k,n-k):
          a[k][j]=last
          last+=1
      for i in range(k+1,n-k):
          a[i][j]=last
          last+=1
      for j in range(n-k-2,k-1,-1):
          a[i][j]=last
          last+=1
      for i in range(n-k-2,k,-1):
          a[i][j]=last
          last+=1

    s=int(math.log(n*n,10))+2 # compute size of cell for printing
    form="{:"+str(s)+"}"
    for i in range(n):
        for j in range(n):
            print(form.format(a[i][j]),end="")
        print("")

sp(3)
# 1 2 3
# 8 9 4
# 7 6 5

sp(4)
#  1  2  3  4
# 12 13 14  5
# 11 16 15  6
# 10  9  8  7

def sp_cc(n): # counterclockwise
    a=[[0 for x in range(n)] for y in range(n)]
    last=1
    for k in range(n//2+1):
      for j in range(n-k-1,k-1,-1):
          a[n-k-1][j]=last
          last+=1
      for i in range(n-k-2,k-1,-1):
          a[i][j]=last
          last+=1
      for j in range(k+1,n-k):
          a[i][j]=last
          last+=1
      for i in range(k+1,n-k-1):
          a[i][j]=last
          last+=1

    s=int(math.log(n*n,10))+2 # compute size of cell for printing
    form="{:"+str(s)+"}"
    for i in range(n):
        for j in range(n):
            print(form.format(a[i][j]),end="")
        print("")

sp_cc(5)
#  9 10 11 12 13
#  8 21 22 23 14
#  7 20 25 24 15
#  6 19 18 17 16
#  5  4  3  2  1

说明

螺旋由同心的正方形组成，例如具有5x5正方形的顺时针旋转看起来像这样：

 5x5        3x3      1x1

>>>>>
^   v       >>>
^   v   +   ^ v   +   >
^   v       <<<
<<<<v

（>>>>>表示“正确执行5次”或将列索引增加5倍，v表示向下或增加行索引，等等。）

所有正方形在大小上都是相同的，我在同心正方形上进行了循环。

对于每个正方形，代码都有四个循环（每侧一个），在每个循环中我们增加或减少列或行索引。如果i是行索引和j列索引，则可以通过以下方式构造5x5的正方形：- j从0 递增到4（5倍）- i从1 递增到4（4倍）- j从3 递减到0（4倍）-递减i 3至1（3次）

对于下一个正方形（3x3和1x1），我们进行了相同的操作，但是适当地移动了初始索引和最终索引。我用了一个索引k为每个同心正方形，有n // 2 + 1个同心正方形。

最后，一些漂亮印刷的数学知识。

要打印索引：

def spi_cc(n): # counter-clockwise
    a=[[0 for x in range(n)] for y in range(n)]
    ind=[]
    last=n*n
    for k in range(n//2+1):
      for j in range(n-k-1,k-1,-1):
          ind.append((n-k-1,j))
      for i in range(n-k-2,k-1,-1):
          ind.append((i,j))
      for j in range(k+1,n-k):
          ind.append((i,j))
      for i in range(k+1,n-k-1):
          ind.append((i,j))

    print(ind)

spi_cc(5)

这是C＃，linq'ish。

public static class SpiralCoords
{
  public static IEnumerable<Tuple<int, int>> GenerateOutTo(int radius)
  {
    //TODO trap negative radius.  0 is ok.

    foreach(int r in Enumerable.Range(0, radius + 1))
    {
      foreach(Tuple<int, int> coord in GenerateRing(r))
      {
        yield return coord;
      }
    }
  }

  public static IEnumerable<Tuple<int, int>> GenerateRing(int radius)
  {
    //TODO trap negative radius.  0 is ok.

    Tuple<int, int> currentPoint = Tuple.Create(radius, 0);
    yield return Tuple.Create(currentPoint.Item1, currentPoint.Item2);

    //move up while we can
    while (currentPoint.Item2 < radius)
    {
      currentPoint.Item2 += 1;
      yield return Tuple.Create(currentPoint.Item1, currentPoint.Item2);
    }
    //move left while we can
    while (-radius < currentPoint.Item1)
    {
      currentPoint.Item1 -=1;
      yield return Tuple.Create(currentPoint.Item1, currentPoint.Item2);    
    }
    //move down while we can
    while (-radius < currentPoint.Item2)
    {
      currentPoint.Item2 -= 1;
      yield return Tuple.Create(currentPoint.Item1, currentPoint.Item2);
    }
    //move right while we can
    while (currentPoint.Item1 < radius)
    {
      currentPoint.Item1 +=1;
      yield return Tuple.Create(currentPoint.Item1, currentPoint.Item2);    
    }
    //move up while we can
    while (currentPoint.Item2 < -1)
    {
      currentPoint.Item2 += 1;
      yield return Tuple.Create(currentPoint.Item1, currentPoint.Item2);
    }
  }

}

该问题的第一个示例（3x3）为：

var coords = SpiralCoords.GenerateOutTo(1);

问题的第二个示例（5x3）为：

var coords = SpiralCoords.GenerateOutTo(2).Where(x => abs(x.Item2) < 2);

这是一个稍微不同的版本-尝试使用recursion和iterators在LUA。在每一步，程序都在矩阵内部进一步下降并循环。我还添加了一个额外的标志螺旋clockwise或anticlockwise。输出从右下角开始，并向中心递归循环。

local row, col, clockwise

local SpiralGen
SpiralGen = function(loop)  -- Generator of elements in one loop
    local startpos = { x = col - loop, y = row - loop }
    local IteratePosImpl = function() -- This function calculates returns the cur, next position in a loop. If called without check, it loops infinitely

        local nextpos = {x = startpos.x, y = startpos.y}        
        local step = clockwise and {x = 0, y = -1} or { x = -1, y = 0 }

        return function()

            curpos = {x = nextpos.x, y = nextpos.y}
            nextpos.x = nextpos.x + step.x
            nextpos.y = nextpos.y + step.y
            if (((nextpos.x == loop or nextpos.x == col - loop + 1) and step.y == 0) or 
                ((nextpos.y == loop or nextpos.y == row - loop + 1) and step.x == 0)) then --Hit a corner in the loop

                local tempstep = {x = step.x, y = step.y}
                step.x = clockwise and tempstep.y or -tempstep.y
                step.y = clockwise and -tempstep.x or tempstep.x
                -- retract next step with new step
                nextpos.x = curpos.x + step.x 
                nextpos.y = curpos.y + step.y

            end         
            return curpos, nextpos
        end
    end
    local IteratePos = IteratePosImpl() -- make an instance
    local curpos, nextpos = IteratePos()
    while (true) do
        if(nextpos.x == startpos.x and nextpos.y == startpos.y) then            
            coroutine.yield(curpos)
            SpiralGen(loop+1) -- Go one step inner, since we're done with this loop
            break -- done with inner loop, get out
        else
            if(curpos.x < loop + 1 or curpos.x > col - loop or curpos.y < loop + 1 or curpos.y > row - loop) then
                break -- done with all elemnts, no place to loop further, break out of recursion
            else
                local curposL = {x = curpos.x, y = curpos.y}
                curpos, nextpos = IteratePos()
                coroutine.yield(curposL)
            end
        end     
    end 
end


local Spiral = function(rowP, colP, clockwiseP)
    row = rowP
    col = colP
    clockwise = clockwiseP
    return coroutine.wrap(function() SpiralGen(0) end) -- make a coroutine that returns all the values as an iterator
end


--test
for pos in Spiral(10,2,true) do
    print (pos.y, pos.x)
end

for pos in Spiral(10,9,false) do
    print (pos.y, pos.x)
end

// PHP实现

function spiral($n) {

    $r = intval((sqrt($n + 1) - 1) / 2) + 1;

    // compute radius : inverse arithmetic sum of 8+16+24+...=
    $p = (8 * $r * ($r - 1)) / 2;
    // compute total point on radius -1 : arithmetic sum of 8+16+24+...

    $en = $r * 2;
    // points by face

    $a = (1 + $n - $p) % ($r * 8);
    // compute de position and shift it so the first is (-r,-r) but (-r+1,-r)
    // so square can connect

    $pos = array(0, 0, $r);
    switch (intval($a / ($r * 2))) {
        // find the face : 0 top, 1 right, 2, bottom, 3 left
        case 0:
            $pos[0] = $a - $r;
            $pos[1] = -$r;
            break;
        case 1:
            $pos[0] = $r;
            $pos[1] = ($a % $en) - $r;
            break;
        case 2:
            $pos[0] = $r - ($a % $en);
            $pos[1] = $r;
            break;
        case 3:
            $pos[0] = -$r;
            $pos[1] = $r - ($a % $en);
            break;
    }
    return $pos;
}

for ($i = 0; $i < 168; $i++) {

    echo '<pre>';
    print_r(spiral($i));
    echo '</pre>';
}

这是此问题的JavaScript（ES6）迭代解决方案：

let spiralMatrix = (x, y, step, count) => {
    let distance = 0;
    let range = 1;
    let direction = 'up';

    for ( let i = 0; i < count; i++ ) {
        console.log('x: '+x+', y: '+y);
        distance++;
        switch ( direction ) {
            case 'up':
                y += step;
                if ( distance >= range ) {
                    direction = 'right';
                    distance = 0;
                }
                break;
            case 'right':
                x += step;
                if ( distance >= range ) {
                    direction = 'bottom';
                    distance = 0;
                    range += 1;
                }
                break;
            case 'bottom':
                y -= step;
                if ( distance >= range ) {
                    direction = 'left';
                    distance = 0;
                }
                break;
            case 'left':
                x -= step;
                if ( distance >= range ) {
                    direction = 'up';
                    distance = 0;
                    range += 1;
                }
                break;
            default:
                break;
        }
    }
}

使用方法如下：

spiralMatrix(0, 0, 1, 100);

这将创建一个向外螺旋，从坐标（x = 0，y = 0）开始，步长为1，总项数等于100。实现始终按以下顺序开始移动-上，右，下，剩下。

请注意，此实现会创建平方矩阵。

这是朱莉娅（Julia）的一个答案：我的方法是在原点周围的同心正方形（“螺旋”）中分配点(0,0)，每个正方形的边长m = 2n + 1，以生成有序字典，其中位置编号（从原点的1开始）为键并将相应的坐标作为值。

由于每个螺旋的最大位置在(n,-n)，因此可以通过简单地从该点向后（即，从右下角开始按m-1单位）进行操作，然后对垂直的三个线段重复进行操作，从而找到其余的点。m-1单元。

此过程在下面以相反的顺序编写，对应于螺旋如何进行而不是在此反向计数过程中进行，即ra[右升]段先递减3(m+1)，然后la[左升] 段再递减2(m+1)，依此类推-希望这是不言而喻的。

import DataStructures: OrderedDict, merge

function spiral(loc::Int)
    s = sqrt(loc-1) |> floor |> Int
    if s % 2 == 0
        s -= 1
    end
    s = (s+1)/2 |> Int
    return s
end

function perimeter(n::Int)
    n > 0 || return OrderedDict([1,[0,0]])
    m = 2n + 1 # width/height of the spiral [square] indexed by n
    # loc_max = m^2
    # loc_min = (2n-1)^2 + 1
    ra = [[m^2-(y+3m-3), [n,n-y]] for y in (m-2):-1:0]
    la = [[m^2-(y+2m-2), [y-n,n]] for y in (m-2):-1:0]
    ld = [[m^2-(y+m-1), [-n,y-n]] for y in (m-2):-1:0]
    rd = [[m^2-y, [n-y,-n]] for y in (m-2):-1:0]
    return OrderedDict(vcat(ra,la,ld,rd))
end

function walk(n)
    cds = OrderedDict(1 => [0,0])
    n > 0 || return cds
    for i in 1:n
        cds = merge(cds, perimeter(i))
    end
    return cds
end

因此，对于您的第一个示例，将其m = 3插入方程式中以找到n给出n = (5-1)/2 = 2，并walk(2)给出坐标位置的有序字典，您可以通过访问字典的vals字段将其转换为坐标数组：

walk(2)
DataStructures.OrderedDict{Any,Any} with 25 entries:
  1  => [0,0]
  2  => [1,0]
  3  => [1,1]
  4  => [0,1]
  ⋮  => ⋮

[(co[1],co[2]) for co in walk(2).vals]
25-element Array{Tuple{Int64,Int64},1}:
 (0,0)  
 (1,0)  
 ⋮       
 (1,-2) 
 (2,-2)

请注意，对于某些函数，例如[eg norm]，最好将坐标留在数组中而不是Tuple{Int,Int}，但是在这里，我将其更改为元组(x,y)-根据要求，使用列表理解。

对于“支持”未规定非方阵（注意，这个解决方案仍然计算离网型值），但上下文，如果你想过滤器只的范围内x通过y（这里x=5，y=3计算全螺后）然后intersect将此矩阵与中的值相对walk。

grid = [[x,y] for x in -2:2, y in -1:1]
5×3 Array{Array{Int64,1},2}:
 [-2,-1]  [-2,0]  [-2,1]
   ⋮       ⋮       ⋮ 
 [2,-1]   [2,0]   [2,1]

[(co[1],co[2]) for co in intersect(walk(2).vals, grid)]
15-element Array{Tuple{Int64,Int64},1}:
 (0,0)  
 (1,0)  
 ⋮ 
 (-2,0) 
 (-2,-1)

您的问题看起来像一个称为螺旋记忆的问题。在这个问题中，网格上的每个正方形都是从位于原点的数字1开始以螺旋状分配的。然后向上计数，同时向外盘旋。例如：

17  16  15  14  13

18   5   4   3  12

19   6   1   2  11

20   7   8   9  10

21  22  23  ---->

我的计算此螺旋模式下每个数字坐标的解决方案发布在下面：

def spiral_pattern(num):
    x = y = 0
    for _ in range(num-1):
        x, y = find_next(x, y)
    yield (x, y)


def find_next(x, y):
    """find the coordinates of the next number"""
    if x == 0 and y == 0:
        return 1, 0

    if abs(x) == abs(y):
        if x > 0 and y > 0:
            x, y = left(x, y)
        elif x < 0 and y > 0:
            x, y = down(x, y)
        elif x < 0 and y < 0:
            x, y = right(x, y)
        elif x > 0 and y < 0:
            x, y = x+1, y
    else:
        if x > y and abs(x) > abs(y):
            x, y = up(x, y)
        elif x < y and abs(x) < abs(y):
            x, y = left(x, y)
        elif x < y and abs(x) > abs(y):
            x, y = down(x, y)
        elif x > y and abs(x) < abs(y):
            x, y = right(x, y)

    return x, y

def up(x, y):
    return x, y+1


def down(x, y):
    return x, y-1


def left(x, y):
    return x-1, y


def right(x, y):
    return x+1, y

这是基于您自己的解决方案，但是我们可以更明智地找到角落。如果M和N非常不同，这将使您更轻松地了解如何跳过外部区域。

def spiral(X, Y):
    x = y = 0
    dx = 0
    dy = -1
    s=0
    ds=2
    for i in range(max(X, Y)**2):
            if abs(x) <= X and abs(y) <= Y/2:
                    print (x, y)
                    # DO STUFF...
            if i==s:
                    dx, dy = -dy, dx
                    s, ds = s+ds/2, ds+1
            x, y = x+dx, y+dy

和优于O（max（n，m）^ 2）的基于生成器的解决方案，它是O（nm + abs（nm）^ 2），因为如果它们不属于解决方案，则会跳过整个条带。

def spiral(X,Y):
X = X+1>>1
Y = Y+1>>1
x = y = 0
d = side = 1
while x<X or y<Y:
    if abs(y)<Y:
        for x in range(x, x+side, d):
            if abs(x)<X: yield x,y
        x += d
    else:
        x += side
    if abs(x)<X:
        for y in range(y, y+side, d):
            if abs(y)<Y: yield x,y
        y += d
    else:
        y += side
    d =-d
    side = d-side

Here is my attempt for simple C solution. First print the outer spiral and move one block inside..and repeat.

#define ROWS        5
#define COLS        5
//int A[ROWS][COLS] = { {1, 2, 3, 4}, {5, 6, 7, 8}, {11, 12, 13, 14}, {15, 16, 17, 18} };
//int A[ROWS][COLS] = { {1, 2, 3}, {6, 7, 8}, { 12, 13, 14} };
//int A[ROWS][COLS] = { {1, 2}, {3, 4}};

int A[ROWS][COLS] = { {1, 2, 3, 4, 5}, {6, 7, 8, 9, 10}, {11, 12, 13, 14, 15} , {16, 17, 18, 19, 20}, {21, 22, 23, 24, 25} };


void print_spiral(int rows, int cols)
{
    int row = 0;
    int offset = 0;

    while (offset < (ROWS - 1)) {
        /* print one outer loop at a time. */
        for (int col = offset; col <= cols; col++) {
            printf("%d ", A[offset][col]);
        }

        for (row = offset + 1; row <= rows; row++) {
            printf("%d ", A[row][cols]);
        }

        for (int col = cols - 1; col >= offset; col--) {
            printf("%d ", A[rows][col]);
        }

        for (row = rows - 1; row >= offset + 1; row--) {
            printf("%d ", A[row][offset]);
        }

       /* Move one block inside */
        offset++;
        rows--;
        cols--;
    }
    printf("\n");
}

int _tmain(int argc, _TCHAR* argv[])
{
    print_spiral(ROWS-1, COLS-1);
    return 0;
}

这是我非常糟糕的解决方案，仅基于Java的最低知识。在这里，我必须将单位螺旋形地放置在一个场上。不能将单元放置在其他单元的顶部或山脉或海洋中。

要清楚。这不是一个好的解决方案。这是一个非常糟糕的解决方案，增加了其他人的乐趣，以嘲笑它可以做得多么糟糕

private void unitPlacementAlgorithm(Position p, Unit u){
    int i = p.getRow();
    int j = p.getColumn();

    int iCounter = 1;
    int jCounter = 0;

    if (getUnitAt(p) == null) {
            unitMap.put(p, u);
    } else {
        iWhileLoop(i, j, iCounter, jCounter, -1, u);
    }

}

private void iWhileLoop(int i, int j, int iCounter, int jCounter, int fortegn, Unit u){
    if(iCounter == 3) {
        for(int k = 0; k < 3; k++) {
            if(k == 2) { //This was added to make the looping stop after 9 units
                System.out.println("There is no more room around the city");
                return; 
            }
            i--;

            if (getUnitAt(new Position(i, j)) == null 
                && !(getTileAt(new Position(i, j)).getTypeString().equals(GameConstants.OCEANS)) 
                && !(getTileAt(new Position(i, j)).getTypeString().equals(GameConstants.MOUNTAINS))) {
                    unitMap.put(new Position(i, j), u);
                    return;
            }
            iCounter--;
        }
    }

    while (iCounter > 0) {
        if (fortegn > 0) {
            i++;
        } else {
            i--;
        }

        if (getUnitAt(new Position(i, j)) == null 
            && !(getTileAt(new Position(i, j)).getTypeString().equals(GameConstants.OCEANS)) 
            && !(getTileAt(new Position(i, j)).getTypeString().equals(GameConstants.MOUNTAINS))) {
                unitMap.put(new Position(i, j), u);
                return;
        }
        iCounter--;
        jCounter++;
    }
    fortegn *= -1;
    jWhileLoop(i, j, iCounter, jCounter, fortegn, u);
}

private void jWhileLoop(int i, int j, int iCounter, int jCounter,
        int fortegn, Unit u) {
    while (jCounter > 0) {
        if (fortegn > 0) {
            j++;
        } else {
            j--;
        }

        if (getUnitAt(new Position(i, j)) == null 
            && !(getTileAt(new Position(i, j)).getTypeString().equals(GameConstants.OCEANS)) 
            && !(getTileAt(new Position(i, j)).getTypeString().equals(GameConstants.MOUNTAINS))) {
                unitMap.put(new Position(i, j), u);
                return;

        }
        jCounter--;
        iCounter++;
        if (jCounter == 0) {
            iCounter++;
        }

    }
    iWhileLoop(i, j, iCounter, jCounter, fortegn, u);
}

对任何可以实际阅读此内容的人都表示感谢

额外的问题：这个“算法”的运行时间是多少？：P

AutoIt解决方案

#include <Math.au3>
#include <Array.au3>

Func SpiralSearch($xMax,$yMax)
    $x = 0
    $y = 0
    $dx = 0
    $dy = -1
    for $i=0 To _max($xMax, $yMax)^2-1 Step 1
        if -$xMax/2 < $x and $x <= $xMax/2 And -$yMax/2 < $y And $y <= $yMax/2 Then
            MsgBox(0, "We are here ", $x & " " & $y)
        EndIf
        if $x == $y or ($x < 0 and $x == -$y) or ($x > 0 and $x == 1-$y) Then
            _ArraySwap ($dx, $dy)
            $dx=-$dx
        EndIf
        $x += $dx
        $y += $dy
    Next
EndFunc

最近，我遇到了类似的挑战，我必须创建一个2D数组并使用螺旋矩阵算法来排序和打印结果。此C＃代码将与N，N 2D数组一起使用。为了清楚起见，它很冗长，可以根据您的需要进行重构。

//CREATE A NEW MATRIX OF SIZE 4 ROWS BY 4 COLUMNS - SCALE MATRIX SIZE HERE
SpiralMatrix SM = new SpiralMatrix(4, 4);
string myData = SM.Read();


public class SpiralMatrix
{
    //LETS BUILD A NEW MATRIX EVERY TIME WE INSTANTIATE OUR CLASS
    public SpiralMatrix(int Rows, int Cols)
    {
        Matrix = new String[Rows, Cols];

        int pos = 1;
        for(int r = 0; r<Rows; r++){
            for (int c = 0; c < Cols; c++)
            {
                //POPULATE THE MATRIX WITH THE CORRECT ROW,COL COORDINATE
                Matrix[r, c] = pos.ToString();
                pos++;
            }
        }
    }

    //READ MATRIX
    public string Read()
    {
        int Row = 0;
        int Col = 0;

        string S = "";
        bool isDone = false;

        //CHECK tO SEE IF POSITION ZERO IS AVAILABLE
        if(PosAvailable(Row, Col)){
            S = ConsumePos(Row, Col);
        }


        //START READING SPIRAL
        //THIS BLOCK READS A FULL CYCLE OF RIGHT,DOWN,LEFT,UP EVERY ITERATION
        while(!isDone)
        {
            bool goNext = false;

            //READ ALL RIGHT SPACES ON THIS PATH PROGRESSION
            while (PosAvailable(Row, Col+1))
            {
                //Is ReadRight Avail
                Col++;
                S += ConsumePos(Row, Col);
                goNext = true;
            }

            //READ ALL DOWN SPACES ON THIS PATH PROGRESSION
            while(PosAvailable(Row+1, Col)){
                //Is ReadDown Avail
                Row++;
                S += ConsumePos(Row, Col);
                goNext = true;
            }

            //READ ALL LEFT SPACES ON THIS PATH PROGRESSION
            while(PosAvailable(Row, Col-1)){
                //Is ReadLeft Avail
                Col--;
                S += ConsumePos(Row, Col);
                goNext = true;
            }

            //READ ALL UP SPACES ON THIS PATH PROGRESSION
            while(PosAvailable(Row-1, Col)){
                //Is ReadUp Avail
                Row--;
                S += ConsumePos(Row, Col);
                goNext = true;
            }

            if(!goNext){
                //DONE - SET EXIT LOOP FLAG
                isDone = true;
            }
        }

        return S;
    }

    //DETERMINE IF THE POSITION IS AVAILABLE
    public bool PosAvailable(int Row, int Col)
    {
        //MAKE SURE WE ARE WITHIN THE BOUNDS OF THE ARRAY
        if (Row < Matrix.GetLength(0) && Row >= 0
            && Col < Matrix.GetLength(1) && Col >= 0)
        {
            //CHECK COORDINATE VALUE
            if (Matrix[Row, Col] != ConsumeChar)
                return true;
            else
                return false;
        }
        else
        {
            //WE ARE OUT OF BOUNDS
            return false;
        }
    }

    public string ConsumePos(int Row, int Col)
    {
        string n = Matrix[Row, Col];
        Matrix[Row, Col] = ConsumeChar;
        return n;
    }

    public string ConsumeChar = "X";
    public string[,] Matrix;
}

我和一位朋友制作了这个，可以根据Javascript将螺旋调整为画布的纵横比。我获得的最佳解决方案是逐像素进行图像演化，填充整个图像。

希望它能帮助一些人。

var width = 150;
var height = 50;

var x = -(width - height)/2;
var y = 0;
var dx = 1;
var dy = 0;
var x_limit = (width - height)/2;
var y_limit = 0;
var counter = 0;

var canvas = document.getElementById("canvas");
var ctx = canvas.getContext('2d');

setInterval(function(){
   if ((-width/2 < x && x <= width/2)  && (-height/2 < y && y <= height/2)) {
       console.log("[ " + x + " , " +  y + " ]");
       ctx.fillStyle = "#FF0000";
       ctx.fillRect(width/2 + x, height/2 - y,1,1);
   }
   if( dx > 0 ){//Dir right
       if(x > x_limit){
           dx = 0;
           dy = 1;
       }
   }
   else if( dy > 0 ){ //Dir up
       if(y > y_limit){
           dx = -1;
           dy = 0;
       }
   }
   else if(dx < 0){ //Dir left
       if(x < (-1 * x_limit)){
           dx = 0;
           dy = -1;
       }
   }
   else if(dy < 0) { //Dir down
       if(y < (-1 * y_limit)){
           dx = 1;
           dy = 0;
           x_limit += 1;
           y_limit += 1;
       }
   }
   counter += 1;
   //alert (counter);
   x += dx;
   y += dy;      
}, 1);

您可以在http://jsfiddle.net/hitbyatruck/c4Kd6/上看到它的运行情况。只要确保在javascript vars和HTML的属性上更改画布的宽度和高度即可。

在Java语言中只是为了好玩：

function spiral(x, y) {
  var iy = ix = 0
    , hr = (x - 1) / 2
    , vr = (y - 1) / 2
    , tt = x * y
    , matrix = []
    , step = 1
    , dx = 1
    , dy = 0;

  while(matrix.length < tt) {

    if((ix <= hr && ix >= (hr * -1)) && (iy <= vr && (iy >= (vr * -1)))) {
      console.log(ix, iy);
      matrix.push([ix, iy]);
    }

    ix += dx;
    iy += dy;

    // check direction
    if(dx !== 0) {
      // increase step
      if(ix === step && iy === (step * -1)) step++;

      // horizontal range reached
      if(ix === step || (ix === step * -1)) {
        dy = (ix === iy)? (dx * -1) : dx;
        dx = 0;  
      }
    } else {
      // vertical range reached
      if(iy === step || (iy === step * -1)) {
        dx = (ix === iy)? (dy * -1) : dy;
        dy = 0;
      }
    }
  }

  return matrix;
}

var sp = spiral(5, 3);

C＃版本也可以处理非正方形大小。

private static Point[] TraverseSpiral(int width, int height) {
    int numElements = width * height + 1;
    Point[] points = new Point[numElements];

    int x = 0;
    int y = 0;
    int dx = 1;
    int dy = 0;
    int xLimit = width - 0;
    int yLimit = height - 1;
    int counter = 0;

    int currentLength = 1;
    while (counter < numElements) {
        points[counter] = new Point(x, y);

        x += dx;
        y += dy;

        currentLength++;
        if (dx > 0) {
            if (currentLength >= xLimit) {
                dx = 0;
                dy = 1;
                xLimit--;
                currentLength = 0;
            }
        } else if (dy > 0) {
            if (currentLength >= yLimit) {
                dx = -1;
                dy = 0;
                yLimit--;
                currentLength = 0;
            }
        } else if (dx < 0) {
            if (currentLength >= xLimit) {
                dx = 0;
                dy = -1;
                xLimit--;
                currentLength = 0;
            }
        } else if (dy < 0) {
            if (currentLength >= yLimit) {
                dx = 1;
                dy = 0;
                yLimit--;
                currentLength = 0;
            }
        }

        counter++;
    }

    Array.Reverse(points);
    return points;
}

我正在共享为其他目的设计的代码。它是关于查找数组元素@螺旋索引“ index”的列号“ X”和行号“ Y”。此函数采用矩阵的宽度“ w”和高度“ h”以及所需的“索引”。当然，此功能可用于产生相同的所需输出。我认为这是最快的方法（因为它跳过细胞而不是扫描细胞）。

    rec BuildSpiralIndex(long w, long h, long index = -1)
    {  
        long count = 0 , x = -1,  y = -1, dir = 1, phase=0, pos = 0,                            length = 0, totallength = 0;
        bool isVertical = false;
        if(index>=(w*h)) return null;

        do 
        {                
            isVertical = (count % 2) != 0;
            length = (isVertical ? h : w) - count/2 - count%2 ;
            totallength += length;
            count++;
        } while(totallength<index);

        count--; w--; h--;
        phase = (count / 4); pos = (count%4);
        x = (pos > 1 ? phase : w - phase);
        y = ((pos == 1 || pos == 2) ? h - phase : phase) + (1 * (pos == 3 ? 1 : 0));
        dir = pos > 1 ? -1 : 1;
        if (isVertical) y -= (totallength - index - 1) * dir;
        else x -= (totallength - index -1) * dir;
        return new rec { X = x, Y = y };
    }

Python使用Can BerkGüderanswer循环顺时针螺旋代码。

def spiral(X, Y):
    x = y = 0
    dx = 0
    dy = 1
    for i in range(max(X, Y)**2):
        if (-X/2 < x <= X/2) and (-Y/2 < y <= Y/2):
            print (x, y)
            # DO STUFF...
        if x == -y or (x < 0 and x == y) or (x > 0 and x-1 == y):
            dx, dy = dy, -dx
        x, y = x+dx, y+dy

Davidont在VB.Net中的出色解决方案

    Public Function Spiral(n As Integer) As RowCol
    ' given n an index in the squared spiral
    ' p the sum of point in inner square
    ' a the position on the current square
    ' n = p + a
    ' starts with row 0 col -1
    Dim r As Integer = CInt(Math.Floor((Math.Sqrt(n + 1) - 1) / 2) + 1)

    ' compute radius : inverse arithmetic sum of 8+16+24+...=
    Dim p As Integer = (8 * r * (r - 1)) \ 2
    ' compute total point on radius -1 : arithmetic sum of 8+16+24+...

    Dim en As Integer = r * 2
    ' points by face

    Dim a As Integer = (1 + n - p) Mod (r * 8)
    ' compute the position and shift it so the first is (-r,-r) but (-r+1,-r)
    ' so square can connect

    Dim row As Integer
    Dim col As Integer

    Select Case Math.Floor(a \ (r * 2))
        ' find the face : 0 top, 1 right, 2, bottom, 3 left
        Case 0
            row = a - r
            col = -r
        Case 1
            row = r
            col = (a Mod en) - r
        Case 2
            row = r - (a Mod en)
            col = r
        Case 3
            row = -r
            col = r - (a Mod en)
    End Select

    Return New RowCol(row, col)
End Function