Sam*_*Sam 3 java sorting mergesort
我正在尝试创建一个合并排序方法,但它继续给出错误的排序.我在哪里进行更改以使其实际排序数组?代码的哪一部分必须有所不同?感谢您的时间.
public static void mergeSort(int[] array, int left, int lHigh, int right, int rHigh) {
int elements = (rHigh - lHigh +1) ;
int[] temp = new int[elements];
int num = left;
while ((left <= lHigh) && (right <= rHigh)){
if (a[left] <= array[right]) {
temp[num] = array[left];
left++;
}
else {
temp[num] = array[right];
right++;
}
num++;
}
while (left <= right){
temp[num] = array[left]; // I'm getting an exception here, and is it because of the num???
left += 1;
num += 1;
}
while (right <= rHigh) {
temp[num] = array[right];
right += 1;
num += 1;
}
for (int i=0; i < elements; i++){
array[rHigh] = temp[rHigh];
rHigh -= 1;
}
Run Code Online (Sandbox Code Playgroud)
编辑:现在mergeSort没有真正排序数字,有人可以告诉我具体是什么?特别是当我打印"测试合并排序"部分时.
Jer*_*ell 19
首先,我假设这是学术性的而不是实际的,因为你没有使用内置的排序功能.话虽如此,这里有一些帮助,让你朝着正确的方向前进:
通常,可以将合并排序视为两种不同的方法:merge()函数将两个排序列表合并为一个排序列表,mergeSort()以递归方式将列表拆分为单个元素列表.由于已经对单个元素列表进行了排序,因此您将所有列表合并为一个大的排序列表.
这是一些副手伪代码:
merge(A, B):
C = empty list
While A and B are not empty:
If the first element of A is smaller than the first element of B:
Remove first element of A.
Add it to the end of C.
Otherwise:
Remove first element of B.
Add it to the end of C.
If A or B still contains elements, add them to the end of C.
mergeSort(A):
if length of A is 1:
return A
Split A into two lists, L and R.
Q = merge(mergeSort(L), mergeSort(R))
return Q
Run Code Online (Sandbox Code Playgroud)
也许这有助于清理你想去的地方.
如果没有,维基百科上总会有MergeSort.
附加:
为了帮助您,以下是您的代码内联的一些注释.
public static void mergeSort(int[] array, int left, int lHigh, int right, int rHigh) {
// what do lHigh and rHigh represent?
int elements = (rHigh - lHigh +1) ;
int[] temp = new int[elements];
int num = left;
// what does this while loop do **conceptually**?
while ((left <= lHigh) && (right <= rHigh)){
if (a[left] <= a[right]) {
// where is 'pos' declared or defined?
temp[pos] = a[left];
// where is leftLow declared or defined? Did you mean 'left' instead?
leftLow ++;
}
else {
temp[num] = a[right];
right ++;
}
num++;
}
// what does this while loop do **conceptually**?
while (left <= right){
// At this point, what is the value of 'num'?
temp[num] = a[left];
left += 1;
num += 1;
}
while (right <= rHigh) {
temp[num] = a[right];
right += 1;
num += 1;
}
// Maybe you meant a[i] = temp[i]?
for (int i=0; i < elements; i++){
// what happens if rHigh is less than elements at this point? Could
// rHigh ever become negative? This would be a runtime error if it did
a[rHigh] = temp[rHigh];
rHigh -= 1;
}
Run Code Online (Sandbox Code Playgroud)
我故意模糊,所以你想一下算法.尝试将自己的注释插入代码中.如果你可以写出概念上发生的事情,那么你可能不需要Stack Overflow :)
我的想法是你没有正确实现这一点.这是因为看起来你只触摸一次数组元素(或只接近一次).这意味着你有一个最糟糕的情况O(N)排序通常至少需要,据O(N * log N)我所知,实际上更简单的合并排序版本O(N^2).
更多:
在合并排序的最简单实现中,我希望在mergeSort()方法中看到某种递归.这是因为合并排序通常是递归定义的.有一些方法可以使用for和while循环迭代地执行此操作,但在您递归获取之前,我绝对不建议将其作为学习工具.
老实说,我建议你在维基百科文章中找到我的伪代码或伪代码,以实现这一点并重新开始使用你的代码.如果你这样做并且它仍然无法正常工作,请在此处发布,我们将帮助您解决问题.
干杯!
最后:
// Precondition: array[left..lHigh] is sorted and array[right...rHigh] is sorted.
// Postcondition: array[left..rHigh] contains the same elements of the above parts, sorted.
public static void mergeSort(int[] array, int left, int lHigh, int right, int rHigh) {
// temp[] needs to be as large as the number of elements you're sorting (not half!)
//int elements = (rHigh - lHigh +1) ;
int elements = rHigh - left;
int[] temp = new int[elements];
// this is your index into the temp array
int num = left;
// now you need to create indices into your two lists
int iL = left;
int iR = right;
// Pseudo code... when you code this, make use of iR, iL, and num!
while( temp is not full ) {
if( left side is all used up ) {
copy rest of right side in.
make sure that at the end of this temp is full so the
while loop quits.
}
else if ( right side is all used up) {
copy rest of left side in.
make sure that at the end of this temp is full so the
while loop quits.
}
else if (array[iL] < array[iR]) { ... }
else if (array[iL] >= array[iR]) { ... }
}
}
Run Code Online (Sandbox Code Playgroud)
| 归档时间: |
|
| 查看次数: |
23827 次 |
| 最近记录: |