mirror of
https://github.com/krahets/hello-algo.git
synced 2024-12-28 05:16:28 +08:00
913 lines
30 KiB
Markdown
913 lines
30 KiB
Markdown
---
|
||
comments: true
|
||
---
|
||
|
||
# 6.2. 哈希冲突
|
||
|
||
上节提到,**通常情况下哈希函数的输入空间远大于输出空间**,因此哈希冲突是不可避免的。例如,输入空间为全体整数,输出空间为数组容量大小,则必然有多个整数映射至同一数组索引。
|
||
|
||
哈希冲突会导致查询结果错误,严重影响哈希表的可用性。为解决该问题,我们可以每当遇到哈希冲突时就进行哈希表扩容,直至冲突消失为止。此方法简单粗暴且有效,但效率太低,因为哈希表扩容需要进行大量的数据搬运与哈希值计算。为了提升效率,我们换一种思路:
|
||
|
||
1. 改良哈希表数据结构,**使得哈希表可以在存在哈希冲突时正常工作**。
|
||
2. 仅在必要时,即当哈希冲突比较严重时,执行扩容操作。
|
||
|
||
哈希表的结构改良方法主要包括链式地址和开放寻址。
|
||
|
||
## 6.2.1. 链式地址
|
||
|
||
在原始哈希表中,每个桶仅能存储一个键值对。「链式地址 Separate Chaining」将单个元素转换为链表,将键值对作为链表节点,将所有发生冲突的键值对都存储在同一链表中。
|
||
|
||
![链式地址哈希表](hash_collision.assets/hash_table_chaining.png)
|
||
|
||
<p align="center"> Fig. 链式地址哈希表 </p>
|
||
|
||
链式地址下,哈希表的操作方法包括:
|
||
|
||
- **查询元素**:输入 `key` ,经过哈希函数得到数组索引,即可访问链表头节点,然后遍历链表并对比 `key` 以查找目标键值对。
|
||
- **添加元素**:先通过哈希函数访问链表头节点,然后将节点(即键值对)添加到链表中。
|
||
- **删除元素**:根据哈希函数的结果访问链表头部,接着遍历链表以查找目标节点,并将其删除。
|
||
|
||
该方法存在一些局限性,包括:
|
||
|
||
- **占用空间增大**,由于链表或二叉树包含节点指针,相比数组更加耗费内存空间;
|
||
- **查询效率降低**,因为需要线性遍历链表来查找对应元素;
|
||
|
||
以下给出了链式地址哈希表的简单实现,需要注意:
|
||
|
||
- 为了使得代码尽量简短,我们使用列表(动态数组)代替链表。换句话说,哈希表(数组)包含多个桶,每个桶都是一个列表。
|
||
- 以下代码实现了哈希表扩容方法。具体来看,当负载因子超过 $0.75$ 时,我们将哈希表扩容至 $2$ 倍。
|
||
|
||
=== "Java"
|
||
|
||
```java title="hash_map_chaining.java"
|
||
/* 键值对 */
|
||
class Pair {
|
||
public int key;
|
||
public String val;
|
||
|
||
public Pair(int key, String val) {
|
||
this.key = key;
|
||
this.val = val;
|
||
}
|
||
}
|
||
|
||
/* 链式地址哈希表 */
|
||
class HashMapChaining {
|
||
int size; // 键值对数量
|
||
int capacity; // 哈希表容量
|
||
double loadThres; // 触发扩容的负载因子阈值
|
||
int extendRatio; // 扩容倍数
|
||
List<List<Pair>> buckets; // 桶数组
|
||
|
||
/* 构造方法 */
|
||
public HashMapChaining() {
|
||
size = 0;
|
||
capacity = 4;
|
||
loadThres = 2 / 3.0;
|
||
extendRatio = 2;
|
||
buckets = new ArrayList<>(capacity);
|
||
for (int i = 0; i < capacity; i++) {
|
||
buckets.add(new ArrayList<>());
|
||
}
|
||
}
|
||
|
||
/* 哈希函数 */
|
||
int hashFunc(int key) {
|
||
return key % capacity;
|
||
}
|
||
|
||
/* 负载因子 */
|
||
double loadFactor() {
|
||
return (double) size / capacity;
|
||
}
|
||
|
||
/* 查询操作 */
|
||
String get(int key) {
|
||
int index = hashFunc(key);
|
||
List<Pair> bucket = buckets.get(index);
|
||
// 遍历桶,若找到 key 则返回对应 val
|
||
for (Pair pair : bucket) {
|
||
if (pair.key == key) {
|
||
return pair.val;
|
||
}
|
||
}
|
||
// 若未找到 key 则返回 null
|
||
return null;
|
||
}
|
||
|
||
/* 添加操作 */
|
||
void put(int key, String val) {
|
||
// 当负载因子超过阈值时,执行扩容
|
||
if (loadFactor() > loadThres) {
|
||
extend();
|
||
}
|
||
int index = hashFunc(key);
|
||
List<Pair> bucket = buckets.get(index);
|
||
// 遍历桶,若遇到指定 key ,则更新对应 val 并返回
|
||
for (Pair pair : bucket) {
|
||
if (pair.key == key) {
|
||
pair.val = val;
|
||
return;
|
||
}
|
||
}
|
||
// 若无该 key ,则将键值对添加至尾部
|
||
Pair pair = new Pair(key, val);
|
||
bucket.add(pair);
|
||
size++;
|
||
}
|
||
|
||
/* 删除操作 */
|
||
void remove(int key) {
|
||
int index = hashFunc(key);
|
||
List<Pair> bucket = buckets.get(index);
|
||
// 遍历桶,从中删除键值对
|
||
for (Pair pair : bucket) {
|
||
if (pair.key == key)
|
||
bucket.remove(pair);
|
||
}
|
||
size--;
|
||
}
|
||
|
||
/* 扩容哈希表 */
|
||
void extend() {
|
||
// 暂存原哈希表
|
||
List<List<Pair>> bucketsTmp = buckets;
|
||
// 初始化扩容后的新哈希表
|
||
capacity *= extendRatio;
|
||
buckets = new ArrayList<>(capacity);
|
||
for (int i = 0; i < capacity; i++) {
|
||
buckets.add(new ArrayList<>());
|
||
}
|
||
size = 0;
|
||
// 将键值对从原哈希表搬运至新哈希表
|
||
for (List<Pair> bucket : bucketsTmp) {
|
||
for (Pair pair : bucket) {
|
||
put(pair.key, pair.val);
|
||
}
|
||
}
|
||
}
|
||
|
||
/* 打印哈希表 */
|
||
void print() {
|
||
for (List<Pair> bucket : buckets) {
|
||
List<String> res = new ArrayList<>();
|
||
for (Pair pair : bucket) {
|
||
res.add(pair.key + " -> " + pair.val);
|
||
}
|
||
System.out.println(res);
|
||
}
|
||
}
|
||
}
|
||
```
|
||
|
||
=== "C++"
|
||
|
||
```cpp title="hash_map_chaining.cpp"
|
||
/* 键值对 */
|
||
struct Pair {
|
||
public:
|
||
int key;
|
||
string val;
|
||
Pair(int key, string val) {
|
||
this->key = key;
|
||
this->val = val;
|
||
}
|
||
};
|
||
|
||
/* 链式地址哈希表 */
|
||
class HashMapChaining {
|
||
private:
|
||
int size; // 键值对数量
|
||
int capacity; // 哈希表容量
|
||
double loadThres; // 触发扩容的负载因子阈值
|
||
int extendRatio; // 扩容倍数
|
||
vector<vector<Pair *>> buckets; // 桶数组
|
||
|
||
public:
|
||
/* 构造方法 */
|
||
HashMapChaining() : size(0), capacity(4), loadThres(2.0 / 3), extendRatio(2) {
|
||
buckets.resize(capacity);
|
||
}
|
||
|
||
/* 哈希函数 */
|
||
int hashFunc(int key) {
|
||
return key % capacity;
|
||
}
|
||
|
||
/* 负载因子 */
|
||
double loadFactor() {
|
||
return (double)size / (double)capacity;
|
||
}
|
||
|
||
/* 查询操作 */
|
||
string get(int key) {
|
||
int index = hashFunc(key);
|
||
// 遍历桶,若找到 key 则返回对应 val
|
||
for (Pair *pair : buckets[index]) {
|
||
if (pair->key == key) {
|
||
return pair->val;
|
||
}
|
||
}
|
||
// 若未找到 key 则返回 nullptr
|
||
return nullptr;
|
||
}
|
||
|
||
/* 添加操作 */
|
||
void put(int key, string val) {
|
||
// 当负载因子超过阈值时,执行扩容
|
||
if (loadFactor() > loadThres) {
|
||
extend();
|
||
}
|
||
int index = hashFunc(key);
|
||
// 遍历桶,若遇到指定 key ,则更新对应 val 并返回
|
||
for (Pair *pair : buckets[index]) {
|
||
if (pair->key == key) {
|
||
pair->val = val;
|
||
return;
|
||
}
|
||
}
|
||
// 若无该 key ,则将键值对添加至尾部
|
||
buckets[index].push_back(new Pair(key, val));
|
||
size++;
|
||
}
|
||
|
||
/* 删除操作 */
|
||
void remove(int key) {
|
||
int index = hashFunc(key);
|
||
auto &bucket = buckets[index];
|
||
// 遍历桶,从中删除键值对
|
||
for (int i = 0; i < bucket.size(); i++) {
|
||
if (bucket[i]->key == key) {
|
||
Pair *tmp = bucket[i];
|
||
bucket.erase(bucket.begin() + i); // 从中删除键值对
|
||
delete tmp; // 释放内存
|
||
size--;
|
||
return;
|
||
}
|
||
}
|
||
}
|
||
|
||
/* 扩容哈希表 */
|
||
void extend() {
|
||
// 暂存原哈希表
|
||
vector<vector<Pair *>> bucketsTmp = buckets;
|
||
// 初始化扩容后的新哈希表
|
||
capacity *= extendRatio;
|
||
buckets.clear();
|
||
buckets.resize(capacity);
|
||
size = 0;
|
||
// 将键值对从原哈希表搬运至新哈希表
|
||
for (auto &bucket : bucketsTmp) {
|
||
for (Pair *pair : bucket) {
|
||
put(pair->key, pair->val);
|
||
}
|
||
}
|
||
}
|
||
|
||
/* 打印哈希表 */
|
||
void print() {
|
||
for (auto &bucket : buckets) {
|
||
cout << "[";
|
||
for (Pair *pair : bucket) {
|
||
cout << pair->key << " -> " << pair->val << ", ";
|
||
}
|
||
cout << "]\n";
|
||
}
|
||
}
|
||
};
|
||
```
|
||
|
||
=== "Python"
|
||
|
||
```python title="hash_map_chaining.py"
|
||
class Pair:
|
||
"""键值对"""
|
||
|
||
def __init__(self, key: int, val: str):
|
||
self.key = key
|
||
self.val = val
|
||
|
||
class HashMapChaining:
|
||
"""链式地址哈希表"""
|
||
|
||
def __init__(self):
|
||
"""构造方法"""
|
||
self.size = 0 # 键值对数量
|
||
self.capacity = 4 # 哈希表容量
|
||
self.load_thres = 2 / 3 # 触发扩容的负载因子阈值
|
||
self.extend_ratio = 2 # 扩容倍数
|
||
self.buckets = [[] for _ in range(self.capacity)] # 桶数组
|
||
|
||
def hash_func(self, key: int) -> int:
|
||
"""哈希函数"""
|
||
return key % self.capacity
|
||
|
||
def load_factor(self) -> float:
|
||
"""负载因子"""
|
||
return self.size / self.capacity
|
||
|
||
def get(self, key: int) -> str:
|
||
"""查询操作"""
|
||
index = self.hash_func(key)
|
||
bucket = self.buckets[index]
|
||
# 遍历桶,若找到 key 则返回对应 val
|
||
for pair in bucket:
|
||
if pair.key == key:
|
||
return pair.val
|
||
# 若未找到 key 则返回 None
|
||
return None
|
||
|
||
def put(self, key: int, val: str):
|
||
"""添加操作"""
|
||
# 当负载因子超过阈值时,执行扩容
|
||
if self.load_factor() > self.load_thres:
|
||
self.extend()
|
||
index = self.hash_func(key)
|
||
bucket = self.buckets[index]
|
||
# 遍历桶,若遇到指定 key ,则更新对应 val 并返回
|
||
for pair in bucket:
|
||
if pair.key == key:
|
||
pair.val = val
|
||
return
|
||
# 若无该 key ,则将键值对添加至尾部
|
||
pair = Pair(key, val)
|
||
bucket.append(pair)
|
||
self.size += 1
|
||
|
||
def remove(self, key: int):
|
||
"""删除操作"""
|
||
index = self.hash_func(key)
|
||
bucket = self.buckets[index]
|
||
# 遍历桶,从中删除键值对
|
||
for pair in bucket:
|
||
if pair.key == key:
|
||
bucket.remove(pair)
|
||
self.size -= 1
|
||
return
|
||
|
||
def extend(self):
|
||
"""扩容哈希表"""
|
||
# 暂存原哈希表
|
||
buckets = self.buckets
|
||
# 初始化扩容后的新哈希表
|
||
self.capacity *= self.extend_ratio
|
||
self.buckets = [[] for _ in range(self.capacity)]
|
||
self.size = 0
|
||
# 将键值对从原哈希表搬运至新哈希表
|
||
for bucket in buckets:
|
||
for pair in bucket:
|
||
self.put(pair.key, pair.val)
|
||
|
||
def print(self):
|
||
"""打印哈希表"""
|
||
for bucket in self.buckets:
|
||
res = []
|
||
for pair in bucket:
|
||
res.append(str(pair.key) + " -> " + pair.val)
|
||
print(res)
|
||
```
|
||
|
||
=== "Go"
|
||
|
||
```go title="hash_map_chaining.go"
|
||
[class]{pair}-[func]{}
|
||
|
||
[class]{hashMapChaining}-[func]{}
|
||
```
|
||
|
||
=== "JavaScript"
|
||
|
||
```javascript title="hash_map_chaining.js"
|
||
[class]{Pair}-[func]{}
|
||
|
||
[class]{HashMapChaining}-[func]{}
|
||
```
|
||
|
||
=== "TypeScript"
|
||
|
||
```typescript title="hash_map_chaining.ts"
|
||
[class]{Pair}-[func]{}
|
||
|
||
[class]{HashMapChaining}-[func]{}
|
||
```
|
||
|
||
=== "C"
|
||
|
||
```c title="hash_map_chaining.c"
|
||
[class]{pair}-[func]{}
|
||
|
||
[class]{hashMapChaining}-[func]{}
|
||
```
|
||
|
||
=== "C#"
|
||
|
||
```csharp title="hash_map_chaining.cs"
|
||
[class]{Pair}-[func]{}
|
||
|
||
[class]{HashMapChaining}-[func]{}
|
||
```
|
||
|
||
=== "Swift"
|
||
|
||
```swift title="hash_map_chaining.swift"
|
||
[class]{Pair}-[func]{}
|
||
|
||
[class]{HashMapChaining}-[func]{}
|
||
```
|
||
|
||
=== "Zig"
|
||
|
||
```zig title="hash_map_chaining.zig"
|
||
[class]{Pair}-[func]{}
|
||
|
||
[class]{HashMapChaining}-[func]{}
|
||
```
|
||
|
||
=== "Dart"
|
||
|
||
```dart title="hash_map_chaining.dart"
|
||
[class]{Pair}-[func]{}
|
||
|
||
[class]{HashMapChaining}-[func]{}
|
||
```
|
||
|
||
!!! tip
|
||
|
||
为了提高效率,**我们可以将链表转换为「AVL 树」或「红黑树」**,从而将查询操作的时间复杂度优化至 $O(\log n)$ 。
|
||
|
||
## 6.2.2. 开放寻址
|
||
|
||
「开放寻址 Open Addressing」不引入额外的数据结构,而是通过“多次探测”来处理哈希冲突,探测方式主要包括线性探测、平方探测、多次哈希。
|
||
|
||
### 线性探测
|
||
|
||
线性探测采用固定步长的线性查找来进行探测,对应的哈希表操作方法为:
|
||
|
||
- **插入元素**:通过哈希函数计算数组索引,若发现桶内已有元素,则从冲突位置向后线性遍历(步长通常为 $1$ ),直至找到空位,将元素插入其中。
|
||
- **查找元素**:若发现哈希冲突,则使用相同步长向后线性遍历,直到找到对应元素,返回 `value` 即可;或者若遇到空位,说明目标键值对不在哈希表中,返回 $\text{None}$ 。
|
||
|
||
![线性探测](hash_collision.assets/hash_table_linear_probing.png)
|
||
|
||
<p align="center"> Fig. 线性探测 </p>
|
||
|
||
然而,线性探测存在以下缺陷:
|
||
|
||
- **不能直接删除元素**。删除元素会在数组内产生一个空位,查找其他元素时,该空位可能导致程序误判元素不存在。因此,需要借助一个标志位来标记已删除元素。
|
||
- **容易产生聚集**。数组内连续被占用位置越长,这些连续位置发生哈希冲突的可能性越大,进一步促使这一位置的“聚堆生长”,最终导致增删查改操作效率降低。
|
||
|
||
如以下代码所示,为开放寻址(线性探测)哈希表的简单实现,重点包括:
|
||
|
||
- 我们使用一个固定的键值对实例 `removed` 来标记已删除元素。也就是说,当一个桶为 $\text{None}$ 或 `removed` 时,这个桶都是空的,可用于放置键值对。
|
||
- 被标记为已删除的空间是可以再次被使用的。当插入元素时,若通过哈希函数找到了被标记为已删除的索引,则可将该元素放置到该索引。
|
||
- 在线性探测时,我们从当前索引 `index` 向后遍历;而当越过数组尾部时,需要回到头部继续遍历。
|
||
|
||
=== "Java"
|
||
|
||
```java title="hash_map_open_addressing.java"
|
||
/* 键值对 */
|
||
class Pair {
|
||
public int key;
|
||
public String val;
|
||
|
||
public Pair(int key, String val) {
|
||
this.key = key;
|
||
this.val = val;
|
||
}
|
||
}
|
||
|
||
/* 开放寻址哈希表 */
|
||
class HashMapOpenAddressing {
|
||
private int size; // 键值对数量
|
||
private int capacity; // 哈希表容量
|
||
private double loadThres; // 触发扩容的负载因子阈值
|
||
private int extendRatio; // 扩容倍数
|
||
private Pair[] buckets; // 桶数组
|
||
private Pair removed; // 删除标记
|
||
|
||
/* 构造方法 */
|
||
public HashMapOpenAddressing() {
|
||
size = 0;
|
||
capacity = 4;
|
||
loadThres = 2.0 / 3.0;
|
||
extendRatio = 2;
|
||
buckets = new Pair[capacity];
|
||
removed = new Pair(-1, "-1");
|
||
}
|
||
|
||
/* 哈希函数 */
|
||
public int hashFunc(int key) {
|
||
return key % capacity;
|
||
}
|
||
|
||
/* 负载因子 */
|
||
public double loadFactor() {
|
||
return (double) size / capacity;
|
||
}
|
||
|
||
/* 查询操作 */
|
||
public String get(int key) {
|
||
int index = hashFunc(key);
|
||
// 线性探测,从 index 开始向后遍历
|
||
for (int i = 0; i < capacity; i++) {
|
||
// 计算桶索引,越过尾部返回头部
|
||
int j = (index + i) % capacity;
|
||
// 若遇到空桶,说明无此 key ,则返回 null
|
||
if (buckets[j] == null)
|
||
return null;
|
||
// 若遇到指定 key ,则返回对应 val
|
||
if (buckets[j].key == key && buckets[j] != removed)
|
||
return buckets[j].val;
|
||
}
|
||
return null;
|
||
}
|
||
|
||
/* 添加操作 */
|
||
public void put(int key, String val) {
|
||
// 当负载因子超过阈值时,执行扩容
|
||
if (loadFactor() > loadThres) {
|
||
extend();
|
||
}
|
||
int index = hashFunc(key);
|
||
// 线性探测,从 index 开始向后遍历
|
||
for (int i = 0; i < capacity; i++) {
|
||
// 计算桶索引,越过尾部返回头部
|
||
int j = (index + i) % capacity;
|
||
// 若遇到空桶、或带有删除标记的桶,则将键值对放入该桶
|
||
if (buckets[j] == null || buckets[j] == removed) {
|
||
buckets[j] = new Pair(key, val);
|
||
size += 1;
|
||
return;
|
||
}
|
||
// 若遇到指定 key ,则更新对应 val
|
||
if (buckets[j].key == key) {
|
||
buckets[j].val = val;
|
||
return;
|
||
}
|
||
}
|
||
}
|
||
|
||
/* 删除操作 */
|
||
public void remove(int key) {
|
||
int index = hashFunc(key);
|
||
// 线性探测,从 index 开始向后遍历
|
||
for (int i = 0; i < capacity; i++) {
|
||
// 计算桶索引,越过尾部返回头部
|
||
int j = (index + i) % capacity;
|
||
// 若遇到空桶,说明无此 key ,则直接返回
|
||
if (buckets[j] == null) {
|
||
return;
|
||
}
|
||
// 若遇到指定 key ,则标记删除并返回
|
||
if (buckets[j].key == key) {
|
||
buckets[j] = removed;
|
||
size -= 1;
|
||
return;
|
||
}
|
||
}
|
||
}
|
||
|
||
/* 扩容哈希表 */
|
||
public void extend() {
|
||
// 暂存原哈希表
|
||
Pair[] bucketsTmp = buckets;
|
||
// 初始化扩容后的新哈希表
|
||
capacity *= extendRatio;
|
||
buckets = new Pair[capacity];
|
||
size = 0;
|
||
// 将键值对从原哈希表搬运至新哈希表
|
||
for (Pair pair : bucketsTmp) {
|
||
if (pair != null && pair != removed) {
|
||
put(pair.key, pair.val);
|
||
}
|
||
}
|
||
}
|
||
|
||
/* 打印哈希表 */
|
||
public void print() {
|
||
for (Pair pair : buckets) {
|
||
if (pair != null) {
|
||
System.out.println(pair.key + " -> " + pair.val);
|
||
} else {
|
||
System.out.println("null");
|
||
}
|
||
}
|
||
}
|
||
}
|
||
```
|
||
|
||
=== "C++"
|
||
|
||
```cpp title="hash_map_open_addressing.cpp"
|
||
/* 键值对 */
|
||
struct Pair {
|
||
int key;
|
||
string val;
|
||
|
||
Pair(int k, string v) : key(k), val(v) {
|
||
}
|
||
};
|
||
|
||
/* 开放寻址哈希表 */
|
||
class HashMapOpenAddressing {
|
||
private:
|
||
int size; // 键值对数量
|
||
int capacity; // 哈希表容量
|
||
double loadThres; // 触发扩容的负载因子阈值
|
||
int extendRatio; // 扩容倍数
|
||
vector<Pair *> buckets; // 桶数组
|
||
Pair *removed; // 删除标记
|
||
|
||
public:
|
||
/* 构造方法 */
|
||
HashMapOpenAddressing() {
|
||
// 构造方法
|
||
size = 0;
|
||
capacity = 4;
|
||
loadThres = 2.0 / 3.0;
|
||
extendRatio = 2;
|
||
buckets = vector<Pair *>(capacity, nullptr);
|
||
removed = new Pair(-1, "-1");
|
||
}
|
||
|
||
/* 哈希函数 */
|
||
int hashFunc(int key) {
|
||
return key % capacity;
|
||
}
|
||
|
||
/* 负载因子 */
|
||
double loadFactor() {
|
||
return static_cast<double>(size) / capacity;
|
||
}
|
||
|
||
/* 查询操作 */
|
||
string get(int key) {
|
||
int index = hashFunc(key);
|
||
// 线性探测,从 index 开始向后遍历
|
||
for (int i = 0; i < capacity; i++) {
|
||
// 计算桶索引,越过尾部返回头部
|
||
int j = (index + i) % capacity;
|
||
// 若遇到空桶,说明无此 key ,则返回 nullptr
|
||
if (buckets[j] == nullptr)
|
||
return nullptr;
|
||
// 若遇到指定 key ,则返回对应 val
|
||
if (buckets[j]->key == key && buckets[j] != removed)
|
||
return buckets[j]->val;
|
||
}
|
||
return nullptr;
|
||
}
|
||
|
||
/* 添加操作 */
|
||
void put(int key, string val) {
|
||
// 当负载因子超过阈值时,执行扩容
|
||
if (loadFactor() > loadThres)
|
||
extend();
|
||
int index = hashFunc(key);
|
||
// 线性探测,从 index 开始向后遍历
|
||
for (int i = 0; i < capacity; i++) {
|
||
// 计算桶索引,越过尾部返回头部
|
||
int j = (index + i) % capacity;
|
||
// 若遇到空桶、或带有删除标记的桶,则将键值对放入该桶
|
||
if (buckets[j] == nullptr || buckets[j] == removed) {
|
||
buckets[j] = new Pair(key, val);
|
||
size += 1;
|
||
return;
|
||
}
|
||
// 若遇到指定 key ,则更新对应 val
|
||
if (buckets[j]->key == key) {
|
||
buckets[j]->val = val;
|
||
return;
|
||
}
|
||
}
|
||
}
|
||
|
||
/* 删除操作 */
|
||
void remove(int key) {
|
||
int index = hashFunc(key);
|
||
// 线性探测,从 index 开始向后遍历
|
||
for (int i = 0; i < capacity; i++) {
|
||
// 计算桶索引,越过尾部返回头部
|
||
int j = (index + i) % capacity;
|
||
// 若遇到空桶,说明无此 key ,则直接返回
|
||
if (buckets[j] == nullptr)
|
||
return;
|
||
// 若遇到指定 key ,则标记删除并返回
|
||
if (buckets[j]->key == key) {
|
||
delete buckets[j]; // 释放内存
|
||
buckets[j] = removed;
|
||
size -= 1;
|
||
return;
|
||
}
|
||
}
|
||
}
|
||
|
||
/* 扩容哈希表 */
|
||
void extend() {
|
||
// 暂存原哈希表
|
||
vector<Pair *> bucketsTmp = buckets;
|
||
// 初始化扩容后的新哈希表
|
||
capacity *= extendRatio;
|
||
buckets = vector<Pair *>(capacity, nullptr);
|
||
size = 0;
|
||
// 将键值对从原哈希表搬运至新哈希表
|
||
for (Pair *pair : bucketsTmp) {
|
||
if (pair != nullptr && pair != removed) {
|
||
put(pair->key, pair->val);
|
||
}
|
||
}
|
||
}
|
||
|
||
/* 打印哈希表 */
|
||
void print() {
|
||
for (auto &pair : buckets) {
|
||
if (pair != nullptr) {
|
||
cout << pair->key << " -> " << pair->val << endl;
|
||
} else {
|
||
cout << "nullptr" << endl;
|
||
}
|
||
}
|
||
}
|
||
};
|
||
```
|
||
|
||
=== "Python"
|
||
|
||
```python title="hash_map_open_addressing.py"
|
||
class Pair:
|
||
"""键值对"""
|
||
|
||
def __init__(self, key: int, val: str):
|
||
self.key = key
|
||
self.val = val
|
||
|
||
class HashMapOpenAddressing:
|
||
"""开放寻址哈希表"""
|
||
|
||
def __init__(self):
|
||
"""构造方法"""
|
||
self.size = 0 # 键值对数量
|
||
self.capacity = 4 # 哈希表容量
|
||
self.load_thres = 2 / 3 # 触发扩容的负载因子阈值
|
||
self.extend_ratio = 2 # 扩容倍数
|
||
self.buckets: list[Pair | None] = [None] * self.capacity # 桶数组
|
||
self.removed = Pair(-1, "-1") # 删除标记
|
||
|
||
def hash_func(self, key: int) -> int:
|
||
"""哈希函数"""
|
||
return key % self.capacity
|
||
|
||
def load_factor(self) -> float:
|
||
"""负载因子"""
|
||
return self.size / self.capacity
|
||
|
||
def get(self, key: int) -> str:
|
||
"""查询操作"""
|
||
index = self.hash_func(key)
|
||
# 线性探测,从 index 开始向后遍历
|
||
for i in range(self.capacity):
|
||
# 计算桶索引,越过尾部返回头部
|
||
j = (index + i) % self.capacity
|
||
# 若遇到空桶,说明无此 key ,则返回 None
|
||
if self.buckets[j] is None:
|
||
return None
|
||
# 若遇到指定 key ,则返回对应 val
|
||
if self.buckets[j].key == key and self.buckets[j] != self.removed:
|
||
return self.buckets[j].val
|
||
|
||
def put(self, key: int, val: str):
|
||
"""添加操作"""
|
||
# 当负载因子超过阈值时,执行扩容
|
||
if self.load_factor() > self.load_thres:
|
||
self.extend()
|
||
index = self.hash_func(key)
|
||
# 线性探测,从 index 开始向后遍历
|
||
for i in range(self.capacity):
|
||
# 计算桶索引,越过尾部返回头部
|
||
j = (index + i) % self.capacity
|
||
# 若遇到空桶、或带有删除标记的桶,则将键值对放入该桶
|
||
if self.buckets[j] in [None, self.removed]:
|
||
self.buckets[j] = Pair(key, val)
|
||
self.size += 1
|
||
return
|
||
# 若遇到指定 key ,则更新对应 val
|
||
if self.buckets[j].key == key:
|
||
self.buckets[j].val = val
|
||
return
|
||
|
||
def remove(self, key: int):
|
||
"""删除操作"""
|
||
index = self.hash_func(key)
|
||
# 线性探测,从 index 开始向后遍历
|
||
for i in range(self.capacity):
|
||
# 计算桶索引,越过尾部返回头部
|
||
j = (index + i) % self.capacity
|
||
# 若遇到空桶,说明无此 key ,则直接返回
|
||
if self.buckets[j] is None:
|
||
return
|
||
# 若遇到指定 key ,则标记删除并返回
|
||
if self.buckets[j].key == key:
|
||
self.buckets[j] = self.removed
|
||
self.size -= 1
|
||
return
|
||
|
||
def extend(self):
|
||
"""扩容哈希表"""
|
||
# 暂存原哈希表
|
||
buckets_tmp = self.buckets
|
||
# 初始化扩容后的新哈希表
|
||
self.capacity *= self.extend_ratio
|
||
self.buckets = [None] * self.capacity
|
||
self.size = 0
|
||
# 将键值对从原哈希表搬运至新哈希表
|
||
for pair in buckets_tmp:
|
||
if pair not in [None, self.removed]:
|
||
self.put(pair.key, pair.val)
|
||
|
||
def print(self) -> None:
|
||
"""打印哈希表"""
|
||
for pair in self.buckets:
|
||
if pair is not None:
|
||
print(pair.key, "->", pair.val)
|
||
else:
|
||
print("None")
|
||
```
|
||
|
||
=== "Go"
|
||
|
||
```go title="hash_map_open_addressing.go"
|
||
[class]{pair}-[func]{}
|
||
|
||
[class]{hashMapOpenAddressing}-[func]{}
|
||
```
|
||
|
||
=== "JavaScript"
|
||
|
||
```javascript title="hash_map_open_addressing.js"
|
||
[class]{Pair}-[func]{}
|
||
|
||
[class]{HashMapOpenAddressing}-[func]{}
|
||
```
|
||
|
||
=== "TypeScript"
|
||
|
||
```typescript title="hash_map_open_addressing.ts"
|
||
[class]{Pair}-[func]{}
|
||
|
||
[class]{HashMapOpenAddressing}-[func]{}
|
||
```
|
||
|
||
=== "C"
|
||
|
||
```c title="hash_map_open_addressing.c"
|
||
[class]{pair}-[func]{}
|
||
|
||
[class]{hashMapOpenAddressing}-[func]{}
|
||
```
|
||
|
||
=== "C#"
|
||
|
||
```csharp title="hash_map_open_addressing.cs"
|
||
[class]{Pair}-[func]{}
|
||
|
||
[class]{HashMapOpenAddressing}-[func]{}
|
||
```
|
||
|
||
=== "Swift"
|
||
|
||
```swift title="hash_map_open_addressing.swift"
|
||
[class]{Pair}-[func]{}
|
||
|
||
[class]{HashMapOpenAddressing}-[func]{}
|
||
```
|
||
|
||
=== "Zig"
|
||
|
||
```zig title="hash_map_open_addressing.zig"
|
||
[class]{Pair}-[func]{}
|
||
|
||
[class]{HashMapOpenAddressing}-[func]{}
|
||
```
|
||
|
||
=== "Dart"
|
||
|
||
```dart title="hash_map_open_addressing.dart"
|
||
[class]{Pair}-[func]{}
|
||
|
||
[class]{HashMapOpenAddressing}-[func]{}
|
||
```
|
||
|
||
### 多次哈希
|
||
|
||
顾名思义,多次哈希方法是使用多个哈希函数 $f_1(x)$ , $f_2(x)$ , $f_3(x)$ , $\cdots$ 进行探测。
|
||
|
||
- **插入元素**:若哈希函数 $f_1(x)$ 出现冲突,则尝试 $f_2(x)$ ,以此类推,直到找到空位后插入元素。
|
||
- **查找元素**:在相同的哈希函数顺序下进行查找,直到找到目标元素时返回;或遇到空位或已尝试所有哈希函数,说明哈希表中不存在该元素,则返回 $\text{None}$ 。
|
||
|
||
与线性探测相比,多次哈希方法不易产生聚集,但多个哈希函数会增加额外的计算量。
|
||
|
||
!!! note "编程语言的选择"
|
||
|
||
Java 采用链式地址。自 JDK 1.8 以来,当 HashMap 内数组长度达到 64 且链表长度达到 8 时,链表会被转换为红黑树以提升查找性能。
|
||
|
||
Python 采用开放寻址。字典 dict 使用伪随机数进行探测。
|
||
|
||
Golang 采用链式地址。Go 规定每个桶最多存储 8 个键值对,超出容量则连接一个溢出桶;当溢出桶过多时,会执行一次特殊的等量扩容操作,以确保性能。
|