Given two words word1 and word2, find the minimum number of operations required to convert word1 to word2.

You have the following 3 operations permitted on a word:

  1. Insert a character
  2. Delete a character
  3. Replace a character

Example 1:

Input: word1 = "horse", word2 = "ros"
Output: 3
Explanation:
horse -> rorse (replace 'h' with 'r')
rorse -> rose (remove 'r')
rose -> ros (remove 'e')

Example 2:

Input: word1 = "intention", word2 = "execution"
Output: 5
Explanation:
intention -> inention (remove 't')
inention -> enention (replace 'i' with 'e')
enention -> exention (replace 'n' with 'x')
exention -> exection (replace 'n' with 'c')
exection -> execution (insert 'u')

这道题让求从一个字符串转变到另一个字符串需要的变换步骤,共有三种变换方式,插入一个字符,删除一个字符,和替换一个字符。题目乍眼一看并不难,但是实际上却暗藏玄机,对于两个字符串的比较,一般都会考虑一下用 HashMap 统计字符出现的频率,但是在这道题却不可以这么做,因为字符串的顺序很重要。还有一种比较常见的错误,就是想当然的认为对于长度不同的两个字符串,长度的差值都是要用插入操作,然后再对应每位字符,不同的地方用修改操作,但是其实这样可能会多用操作,因为删除操作有时同时可以达到修改的效果。比如题目中的例子1,当把 horse 变为 rorse 之后,之后只要删除第二个r,跟最后一个e,就可以变为 ros。实际上只要三步就完成了,因为删除了某个字母后,原来左右不相连的字母现在就连一起了,有可能刚好组成了需要的字符串。所以在比较的时候,要尝试三种操作,因为谁也不知道当前的操作会对后面产生什么样的影响。对于当前比较的两个字符 word1[i] 和 word2[j],若二者相同,一切好说,直接跳到下一个位置。若不相同,有三种处理方法,首先是直接插入一个 word2[j],那么 word2[j] 位置的字符就跳过了,接着比较 word1[i] 和 word2[j+1] 即可。第二个种方法是删除,即将 word1[i] 字符直接删掉,接着比较 word1[i+1] 和 word2[j] 即可。第三种则是将 word1[i] 修改为 word2[j],接着比较 word1[i+1] 和 word[j+1] 即可。分析到这里,就可以直接写出递归的代码,但是很可惜会 Time Limited Exceed,所以必须要优化时间复杂度,需要去掉大量的重复计算,这里使用记忆数组 memo 来保存计算过的状态,从而可以通过 OJ,注意这里的 insertCnt,deleteCnt,replaceCnt 仅仅是表示当前对应的位置分别采用了插入,删除,和替换操作,整体返回的最小距离,后面位置的还是会调用递归返回最小的,参见代码如下:

解法一:

class Solution {
public:
int minDistance(string word1, string word2) {
int m = word1.size(), n = word2.size();
vector<vector<int>> memo(m, vector<int>(n));
return helper(word1, , word2, , memo);
}
int helper(string& word1, int i, string& word2, int j, vector<vector<int>>& memo) {
if (i == word1.size()) return (int)word2.size() - j;
if (j == word2.size()) return (int)word1.size() - i;
if (memo[i][j] > ) return memo[i][j];
int res = ;
if (word1[i] == word2[j]) {
return helper(word1, i + , word2, j + , memo);
} else {
int insertCnt = helper(word1, i, word2, j + , memo);
int deleteCnt = helper(word1, i + , word2, j, memo);
int replaceCnt = helper(word1, i + , word2, j + , memo);
res = min(insertCnt, min(deleteCnt, replaceCnt)) + ;
}
return memo[i][j] = res;
}
};

根据以往的经验,对于字符串相关的题目且求极值的问题,十有八九都是用动态规划 Dynamic Programming 来解,这道题也不例外。其实解法一的递归加记忆数组的方法也可以看作是 DP 的递归写法。这里需要维护一个二维的数组 dp,其大小为 mxn,m和n分别为 word1 和 word2 的长度。dp[i][j] 表示从 word1 的前i个字符转换到 word2 的前j个字符所需要的步骤。先给这个二维数组 dp 的第一行第一列赋值,这个很简单,因为第一行和第一列对应的总有一个字符串是空串,于是转换步骤完全是另一个字符串的长度。跟以往的 DP 题目类似,难点还是在于找出状态转移方程,可以举个例子来看,比如 word1 是 "bbc",word2 是 "abcd",可以得到 dp 数组如下:

  Ø a b c d
Ø
b
b
c

通过观察可以发现,当 word1[i] == word2[j] 时,dp[i][j] = dp[i - 1][j - 1],其他情况时,dp[i][j] 是其左,左上,上的三个值中的最小值加1,其实这里的左,上,和左上,分别对应的增加,删除,修改操作,具体可以参见解法一种的讲解部分,那么可以得到状态转移方程为:

dp[i][j] =      /    dp[i - 1][j - 1]                                                                   if word1[i - 1] == word2[j - 1]

\    min(dp[i - 1][j - 1], min(dp[i - 1][j], dp[i][j - 1])) + 1            else

解法二:

class Solution {
public:
int minDistance(string word1, string word2) {
int m = word1.size(), n = word2.size();
vector<vector<int>> dp(m + , vector<int>(n + ));
for (int i = ; i <= m; ++i) dp[i][] = i;
for (int i = ; i <= n; ++i) dp[][i] = i;
for (int i = ; i <= m; ++i) {
for (int j = ; j <= n; ++j) {
if (word1[i - ] == word2[j - ]) {
dp[i][j] = dp[i - ][j - ];
} else {
dp[i][j] = min(dp[i - ][j - ], min(dp[i - ][j], dp[i][j - ])) + ;
}
}
}
return dp[m][n];
}
};

Github 同步地址:

https://github.com/grandyang/leetcode/issues/72

类似题目:

One Edit Distance

Delete Operation for Two Strings

Minimum ASCII Delete Sum for Two Strings

参考资料:

https://leetcode.com/problems/edit-distance/

https://leetcode.com/problems/edit-distance/discuss/25846/C%2B%2B-O(n)-space-DP

LeetCode All in One 题目讲解汇总(持续更新中...)

[LeetCode] Edit Distance 编辑距离的更多相关文章

  1. [LeetCode] 72. Edit Distance 编辑距离

    Given two words word1 and word2, find the minimum number of operations required to convert word1 to  ...

  2. leetCode 72.Edit Distance (编辑距离) 解题思路和方法

    Edit Distance Given two words word1 and word2, find the minimum number of steps required to convert  ...

  3. Edit Distance编辑距离(NM tag)- sam/bam格式解读进阶

    sam格式很精炼,几乎包含了比对的所有信息,我们平常用到的信息很少,但特殊情况下,我们会用到一些较为生僻的信息,关于这些信息sam官方文档的介绍比较精简,直接看估计很难看懂. 今天要介绍的是如何通过b ...

  4. Leetcode:Edit Distance 解题报告

    Edit Distance Given two words word1 and word2, find the minimum number of steps required to convert  ...

  5. 【LeetCode】72. Edit Distance 编辑距离(Python & C++)

    作者: 负雪明烛 id: fuxuemingzhu 个人博客: http://fuxuemingzhu.cn/ 目录 题目描述 题目大意 解题方法 递归 记忆化搜索 动态规划 日期 题目地址:http ...

  6. 【LeetCode每天一题】Edit Distance(编辑距离)

    Given two words word1 and word2, find the minimum number of operations required to convert word1 to  ...

  7. [leetcode]Edit Distance @ Python

    原题地址:https://oj.leetcode.com/problems/edit-distance/ 题意: Given two words word1 and word2, find the m ...

  8. leetcode72. Edit Distance(编辑距离)

    以下为个人翻译方便理解 编辑距离问题是一个经典的动态规划问题.首先定义dp[i][j表示word1[0..i-1]到word2[0..j-1]的最小操作数(即编辑距离). 状态转换方程有两种情况:边界 ...

  9. 072 Edit Distance 编辑距离

    给出两个单词 word1 和 word2,找出将 word1 转换成 word2 所使用的最少的步骤数 (每个操作记为一步).你可以对一个单词进行以下三种操作:a) 插入一个字符b) 删除一个字符c) ...

随机推荐

  1. 谈谈枚举的新用法——java

    问题的由来 前段时间改游戏buff功能,干了一件愚蠢的事情,那就是把枚举和运算集合在一起,然后运行一段时间后buff就出现各种问题,我当时懵逼了! 事情是这样的,做过游戏的都知道,buff,需要分类型 ...

  2. Rafy 框架 - 大批量导入实体

    某些场景下,开发者希望能够大批量地把实体的数据导入到数据库中.虽然使用实体仓库保存实体列表非常方便,但是其内部实现机制是一条一条的保存到数据库,当实体的个数较多时,效率就会很低.所以 Rafy 设计了 ...

  3. 设计模式(十四)模板方法模式(Template Pattern)

    一.引言 提到模板,大家肯定不免想到生活中的“简历模板”.“论文模板”.“Word中模版文件”等,在现实生活中,模板的概念就是——有一个规定的格式,然后每个人都可以根据自己的需求或情况去更新它,例如简 ...

  4. Servlet的生命周期及工作原理

    Servlet生命周期分为三个阶段: 1,初始化阶段  调用init()方法 2,响应客户请求阶段 调用service()方法 3,终止阶段 调用destroy()方法 Servlet初始化阶段: 在 ...

  5. C标准头文件<ctype.h>

    主要包括了一些字符识别和转换函数 字符判断 isalnum() //函数原型 #include<ctype.h> int isalum(int c); 功能:如果输入的字符是字母(alph ...

  6. DbMigration使用方法

    1.Enable-Migrations -ContextTypeNameLITCS.Data.gmisContext Enable-Migrations  命令创建了一个新的Migrations文件夹 ...

  7. Struts 2 数据校验要用到的类和两种校验方式以及一些校验问题的解决

    通过继承ActionSupport类来完成Action开发,ActionSupport类不仅对Action接口进行简单实现, 同时增加了验证.本地化等支持 .真实开发中自定义Action都需要继承该类 ...

  8. JavaScript闭包(Closure)

    JavaScript闭包(Closure) 本文收集了多本书里对JavaScript闭包(Closure)的解释,或许会对理解闭包有一定帮助. <你不知道的JavsScript> Java ...

  9. 用Kotlin语言重新编写Plaid APP:经验教训(II)

    原文标题:Converting Plaid to Kotlin: Lessons learned (Part 2) 原文链接:http://antonioleiva.com/plaid-kotlin- ...

  10. React-Native坑:Invariant Violation:Application 项目名 has not been registered.

    前言 在学习一门新技术的你也许有跟我一样的困惑,照着书上或者视频上的敲了.但是就是有各种问题没有出来自己想要的结果.我会将自己在这个过程中遇到的坑都记录下来,不一定全覆盖,但希望这些文章可以解决你的问 ...