二分法
前提:二分法查找适用于数据量较大时,但是数据需要先排好顺序(有序数组)。
主要思想是:(设查找的数组区间为array[front, end])
(1)确定该区间的中间位置mid
(2)将查找的值T与array[mid]比较。若相等,查找成功返回此位置mid;
否则确定新的查找区域,继续二分查找。
区域确定如下:
- array[mid]>T 由数组的有序性可知array[mid,mid+1,……,high]>T;
- 故新的区间为array[low,……,mid-1];
- array[mid]<T 类似上面查找区间为array[mid+1,……,high];
- 每一次查找与中间值比较,可以确定是否查找成功,不成功当前查找区间将缩小一半,递归查找即可。
时间复杂度
- 最坏情况查找最后一个元素(或者第一个元素)Master定理T(n)=T(n/2)+O(1)所以T(n)=O(log2n)
- 最好情况查找中间元素O(1)查找的元素即为中间元素(奇数长度数列的正中间,偶数长度数列的中间靠左的元素)
python实现-方法1(递归):
array = [1, 3, 4, 6, 8, 10, 14]
front = 0
end = len(array) -1
def Binary1(array, front, end, t):
if front <= end:
mid = int((front + end) / 2)
print("front = {}, end = {}, mid = {}".format(front, end, mid))
if array[mid] == t:
return mid
elif array[mid] > t:
return Binary1(array, front, mid-1, t)
else:
return Binary1(array, mid+1, end, t)
else:
return -1
Binary1(array, front, end, 10)
front = 0, end = 6, mid = 3
front = 4, end = 6, mid = 5
out: 5
python实现-方法2:
def Binary2(array, front, end, t):
while front <= end:
mid = int((front + end) / 2)
print("front = {}, end = {}, mid = {}".format(front, end, mid))
if array[mid] == t:
return mid
elif array[mid] > t:
end = mid - 1
else:
front = mid + 1
else:
return -1
Binary2(array, front, end, 15)
front = 0, end = 6, mid = 3
front = 4, end = 6, mid = 5
front = 6, end = 6, mid = 6
out: -1