mirror of
https://github.com/krahets/hello-algo.git
synced 2024-12-27 11:26:28 +08:00
1727 lines
41 KiB
Markdown
Executable file
1727 lines
41 KiB
Markdown
Executable file
# 时间复杂度
|
||
|
||
## 统计算法运行时间
|
||
|
||
运行时间可以直观且准确地反映算法的效率。然而,如果我们想要准确预估一段代码的运行时间,应该如何操作呢?
|
||
|
||
1. **确定运行平台**,包括硬件配置、编程语言、系统环境等,这些因素都会影响代码的运行效率。
|
||
2. **评估各种计算操作所需的运行时间**,例如加法操作 `+` 需要 1 ns,乘法操作 `*` 需要 10 ns,打印操作需要 5 ns 等。
|
||
3. **统计代码中所有的计算操作**,并将所有操作的执行时间求和,从而得到运行时间。
|
||
|
||
例如以下代码,输入数据大小为 $n$ ,根据以上方法,可以得到算法运行时间为 $6n + 12$ ns 。
|
||
|
||
$$
|
||
1 + 1 + 10 + (1 + 5) \times n = 6n + 12
|
||
$$
|
||
|
||
=== "Java"
|
||
|
||
```java title=""
|
||
// 在某运行平台下
|
||
void algorithm(int n) {
|
||
int a = 2; // 1 ns
|
||
a = a + 1; // 1 ns
|
||
a = a * 2; // 10 ns
|
||
// 循环 n 次
|
||
for (int i = 0; i < n; i++) { // 1 ns ,每轮都要执行 i++
|
||
System.out.println(0); // 5 ns
|
||
}
|
||
}
|
||
```
|
||
|
||
=== "C++"
|
||
|
||
```cpp title=""
|
||
// 在某运行平台下
|
||
void algorithm(int n) {
|
||
int a = 2; // 1 ns
|
||
a = a + 1; // 1 ns
|
||
a = a * 2; // 10 ns
|
||
// 循环 n 次
|
||
for (int i = 0; i < n; i++) { // 1 ns ,每轮都要执行 i++
|
||
cout << 0 << endl; // 5 ns
|
||
}
|
||
}
|
||
```
|
||
|
||
=== "Python"
|
||
|
||
```python title=""
|
||
# 在某运行平台下
|
||
def algorithm(n: int) -> None:
|
||
a = 2 # 1 ns
|
||
a = a + 1 # 1 ns
|
||
a = a * 2 # 10 ns
|
||
# 循环 n 次
|
||
for _ in range(n): # 1 ns
|
||
print(0) # 5 ns
|
||
```
|
||
|
||
=== "Go"
|
||
|
||
```go title=""
|
||
// 在某运行平台下
|
||
func algorithm(n int) {
|
||
a := 2 // 1 ns
|
||
a = a + 1 // 1 ns
|
||
a = a * 2 // 10 ns
|
||
// 循环 n 次
|
||
for i := 0; i < n; i++ { // 1 ns
|
||
fmt.Println(a) // 5 ns
|
||
}
|
||
}
|
||
```
|
||
|
||
=== "JavaScript"
|
||
|
||
```javascript title=""
|
||
// 在某运行平台下
|
||
function algorithm(n) {
|
||
var a = 2; // 1 ns
|
||
a = a + 1; // 1 ns
|
||
a = a * 2; // 10 ns
|
||
// 循环 n 次
|
||
for(let i = 0; i < n; i++) { // 1 ns ,每轮都要执行 i++
|
||
console.log(0); // 5 ns
|
||
}
|
||
}
|
||
```
|
||
|
||
=== "TypeScript"
|
||
|
||
```typescript title=""
|
||
// 在某运行平台下
|
||
function algorithm(n: number): void {
|
||
var a: number = 2; // 1 ns
|
||
a = a + 1; // 1 ns
|
||
a = a * 2; // 10 ns
|
||
// 循环 n 次
|
||
for(let i = 0; i < n; i++) { // 1 ns ,每轮都要执行 i++
|
||
console.log(0); // 5 ns
|
||
}
|
||
}
|
||
```
|
||
|
||
=== "C"
|
||
|
||
```c title=""
|
||
// 在某运行平台下
|
||
void algorithm(int n) {
|
||
int a = 2; // 1 ns
|
||
a = a + 1; // 1 ns
|
||
a = a * 2; // 10 ns
|
||
// 循环 n 次
|
||
for (int i = 0; i < n; i++) { // 1 ns ,每轮都要执行 i++
|
||
printf("%d", 0); // 5 ns
|
||
}
|
||
}
|
||
```
|
||
|
||
=== "C#"
|
||
|
||
```csharp title=""
|
||
// 在某运行平台下
|
||
void algorithm(int n)
|
||
{
|
||
int a = 2; // 1 ns
|
||
a = a + 1; // 1 ns
|
||
a = a * 2; // 10 ns
|
||
// 循环 n 次
|
||
for (int i = 0; i < n; i++)
|
||
{ // 1 ns ,每轮都要执行 i++
|
||
Console.WriteLine(0); // 5 ns
|
||
}
|
||
}
|
||
```
|
||
|
||
=== "Swift"
|
||
|
||
```swift title=""
|
||
// 在某运行平台下
|
||
func algorithm(n: Int) {
|
||
var a = 2 // 1 ns
|
||
a = a + 1 // 1 ns
|
||
a = a * 2 // 10 ns
|
||
// 循环 n 次
|
||
for _ in 0 ..< n { // 1 ns
|
||
print(0) // 5 ns
|
||
}
|
||
}
|
||
```
|
||
|
||
=== "Zig"
|
||
|
||
```zig title=""
|
||
|
||
```
|
||
|
||
=== "Dart"
|
||
|
||
```dart title=""
|
||
|
||
```
|
||
|
||
然而实际上,**统计算法的运行时间既不合理也不现实**。首先,我们不希望预估时间和运行平台绑定,因为算法需要在各种不同的平台上运行。其次,我们很难获知每种操作的运行时间,这给预估过程带来了极大的难度。
|
||
|
||
## 统计时间增长趋势
|
||
|
||
「时间复杂度分析」采取了一种不同的方法,其统计的不是算法运行时间,**而是算法运行时间随着数据量变大时的增长趋势**。
|
||
|
||
“时间增长趋势”这个概念较为抽象,我们通过一个例子来加以理解。假设输入数据大小为 $n$ ,给定三个算法 `A` , `B` , `C` 。
|
||
|
||
- 算法 `A` 只有 $1$ 个打印操作,算法运行时间不随着 $n$ 增大而增长。我们称此算法的时间复杂度为「常数阶」。
|
||
- 算法 `B` 中的打印操作需要循环 $n$ 次,算法运行时间随着 $n$ 增大呈线性增长。此算法的时间复杂度被称为「线性阶」。
|
||
- 算法 `C` 中的打印操作需要循环 $1000000$ 次,但运行时间仍与输入数据大小 $n$ 无关。因此 `C` 的时间复杂度和 `A` 相同,仍为「常数阶」。
|
||
|
||
=== "Java"
|
||
|
||
```java title=""
|
||
// 算法 A 时间复杂度:常数阶
|
||
void algorithm_A(int n) {
|
||
System.out.println(0);
|
||
}
|
||
// 算法 B 时间复杂度:线性阶
|
||
void algorithm_B(int n) {
|
||
for (int i = 0; i < n; i++) {
|
||
System.out.println(0);
|
||
}
|
||
}
|
||
// 算法 C 时间复杂度:常数阶
|
||
void algorithm_C(int n) {
|
||
for (int i = 0; i < 1000000; i++) {
|
||
System.out.println(0);
|
||
}
|
||
}
|
||
```
|
||
|
||
=== "C++"
|
||
|
||
```cpp title=""
|
||
// 算法 A 时间复杂度:常数阶
|
||
void algorithm_A(int n) {
|
||
cout << 0 << endl;
|
||
}
|
||
// 算法 B 时间复杂度:线性阶
|
||
void algorithm_B(int n) {
|
||
for (int i = 0; i < n; i++) {
|
||
cout << 0 << endl;
|
||
}
|
||
}
|
||
// 算法 C 时间复杂度:常数阶
|
||
void algorithm_C(int n) {
|
||
for (int i = 0; i < 1000000; i++) {
|
||
cout << 0 << endl;
|
||
}
|
||
}
|
||
```
|
||
|
||
=== "Python"
|
||
|
||
```python title=""
|
||
# 算法 A 时间复杂度:常数阶
|
||
def algorithm_A(n: int) -> None:
|
||
print(0)
|
||
# 算法 B 时间复杂度:线性阶
|
||
def algorithm_B(n: int) -> None:
|
||
for _ in range(n):
|
||
print(0)
|
||
# 算法 C 时间复杂度:常数阶
|
||
def algorithm_C(n: int) -> None:
|
||
for _ in range(1000000):
|
||
print(0)
|
||
```
|
||
|
||
=== "Go"
|
||
|
||
```go title=""
|
||
// 算法 A 时间复杂度:常数阶
|
||
func algorithm_A(n int) {
|
||
fmt.Println(0)
|
||
}
|
||
// 算法 B 时间复杂度:线性阶
|
||
func algorithm_B(n int) {
|
||
for i := 0; i < n; i++ {
|
||
fmt.Println(0)
|
||
}
|
||
}
|
||
// 算法 C 时间复杂度:常数阶
|
||
func algorithm_C(n int) {
|
||
for i := 0; i < 1000000; i++ {
|
||
fmt.Println(0)
|
||
}
|
||
}
|
||
```
|
||
|
||
=== "JavaScript"
|
||
|
||
```javascript title=""
|
||
// 算法 A 时间复杂度:常数阶
|
||
function algorithm_A(n) {
|
||
console.log(0);
|
||
}
|
||
// 算法 B 时间复杂度:线性阶
|
||
function algorithm_B(n) {
|
||
for (let i = 0; i < n; i++) {
|
||
console.log(0);
|
||
}
|
||
}
|
||
// 算法 C 时间复杂度:常数阶
|
||
function algorithm_C(n) {
|
||
for (let i = 0; i < 1000000; i++) {
|
||
console.log(0);
|
||
}
|
||
}
|
||
|
||
```
|
||
|
||
=== "TypeScript"
|
||
|
||
```typescript title=""
|
||
// 算法 A 时间复杂度:常数阶
|
||
function algorithm_A(n: number): void {
|
||
console.log(0);
|
||
}
|
||
// 算法 B 时间复杂度:线性阶
|
||
function algorithm_B(n: number): void {
|
||
for (let i = 0; i < n; i++) {
|
||
console.log(0);
|
||
}
|
||
}
|
||
// 算法 C 时间复杂度:常数阶
|
||
function algorithm_C(n: number): void {
|
||
for (let i = 0; i < 1000000; i++) {
|
||
console.log(0);
|
||
}
|
||
}
|
||
```
|
||
|
||
=== "C"
|
||
|
||
```c title=""
|
||
// 算法 A 时间复杂度:常数阶
|
||
void algorithm_A(int n) {
|
||
printf("%d", 0);
|
||
}
|
||
// 算法 B 时间复杂度:线性阶
|
||
void algorithm_B(int n) {
|
||
for (int i = 0; i < n; i++) {
|
||
printf("%d", 0);
|
||
}
|
||
}
|
||
// 算法 C 时间复杂度:常数阶
|
||
void algorithm_C(int n) {
|
||
for (int i = 0; i < 1000000; i++) {
|
||
printf("%d", 0);
|
||
}
|
||
}
|
||
```
|
||
|
||
=== "C#"
|
||
|
||
```csharp title=""
|
||
// 算法 A 时间复杂度:常数阶
|
||
void algorithm_A(int n)
|
||
{
|
||
Console.WriteLine(0);
|
||
}
|
||
// 算法 B 时间复杂度:线性阶
|
||
void algorithm_B(int n)
|
||
{
|
||
for (int i = 0; i < n; i++)
|
||
{
|
||
Console.WriteLine(0);
|
||
}
|
||
}
|
||
// 算法 C 时间复杂度:常数阶
|
||
void algorithm_C(int n)
|
||
{
|
||
for (int i = 0; i < 1000000; i++)
|
||
{
|
||
Console.WriteLine(0);
|
||
}
|
||
}
|
||
```
|
||
|
||
=== "Swift"
|
||
|
||
```swift title=""
|
||
// 算法 A 时间复杂度:常数阶
|
||
func algorithmA(n: Int) {
|
||
print(0)
|
||
}
|
||
|
||
// 算法 B 时间复杂度:线性阶
|
||
func algorithmB(n: Int) {
|
||
for _ in 0 ..< n {
|
||
print(0)
|
||
}
|
||
}
|
||
|
||
// 算法 C 时间复杂度:常数阶
|
||
func algorithmC(n: Int) {
|
||
for _ in 0 ..< 1000000 {
|
||
print(0)
|
||
}
|
||
}
|
||
```
|
||
|
||
=== "Zig"
|
||
|
||
```zig title=""
|
||
|
||
```
|
||
|
||
=== "Dart"
|
||
|
||
```dart title=""
|
||
|
||
```
|
||
|
||
![算法 A, B, C 的时间增长趋势](time_complexity.assets/time_complexity_simple_example.png)
|
||
|
||
相较于直接统计算法运行时间,时间复杂度分析有哪些优势和局限性呢?
|
||
|
||
**时间复杂度能够有效评估算法效率**。例如,算法 `B` 的运行时间呈线性增长,在 $n > 1$ 时比算法 `A` 慢,在 $n > 1000000$ 时比算法 `C` 慢。事实上,只要输入数据大小 $n$ 足够大,复杂度为「常数阶」的算法一定优于「线性阶」的算法,这正是时间增长趋势所表达的含义。
|
||
|
||
**时间复杂度的推算方法更简便**。显然,运行平台和计算操作类型都与算法运行时间的增长趋势无关。因此在时间复杂度分析中,我们可以简单地将所有计算操作的执行时间视为相同的“单位时间”,从而将“计算操作的运行时间的统计”简化为“计算操作的数量的统计”,这样的简化方法大大降低了估算难度。
|
||
|
||
**时间复杂度也存在一定的局限性**。例如,尽管算法 `A` 和 `C` 的时间复杂度相同,但实际运行时间差别很大。同样,尽管算法 `B` 的时间复杂度比 `C` 高,但在输入数据大小 $n$ 较小时,算法 `B` 明显优于算法 `C` 。在这些情况下,我们很难仅凭时间复杂度判断算法效率高低。当然,尽管存在上述问题,复杂度分析仍然是评判算法效率最有效且常用的方法。
|
||
|
||
## 函数渐近上界
|
||
|
||
设算法的计算操作数量是一个关于输入数据大小 $n$ 的函数,记为 $T(n)$ ,则以下算法的操作数量为
|
||
|
||
$$
|
||
T(n) = 3 + 2n
|
||
$$
|
||
|
||
=== "Java"
|
||
|
||
```java title=""
|
||
void algorithm(int n) {
|
||
int a = 1; // +1
|
||
a = a + 1; // +1
|
||
a = a * 2; // +1
|
||
// 循环 n 次
|
||
for (int i = 0; i < n; i++) { // +1(每轮都执行 i ++)
|
||
System.out.println(0); // +1
|
||
}
|
||
}
|
||
```
|
||
|
||
=== "C++"
|
||
|
||
```cpp title=""
|
||
void algorithm(int n) {
|
||
int a = 1; // +1
|
||
a = a + 1; // +1
|
||
a = a * 2; // +1
|
||
// 循环 n 次
|
||
for (int i = 0; i < n; i++) { // +1(每轮都执行 i ++)
|
||
cout << 0 << endl; // +1
|
||
}
|
||
}
|
||
```
|
||
|
||
=== "Python"
|
||
|
||
```python title=""
|
||
def algorithm(n: int) -> None:
|
||
a = 1 # +1
|
||
a = a + 1 # +1
|
||
a = a * 2 # +1
|
||
# 循环 n 次
|
||
for i in range(n): # +1
|
||
print(0) # +1
|
||
```
|
||
|
||
=== "Go"
|
||
|
||
```go title=""
|
||
func algorithm(n int) {
|
||
a := 1 // +1
|
||
a = a + 1 // +1
|
||
a = a * 2 // +1
|
||
// 循环 n 次
|
||
for i := 0; i < n; i++ { // +1
|
||
fmt.Println(a) // +1
|
||
}
|
||
}
|
||
```
|
||
|
||
=== "JavaScript"
|
||
|
||
```javascript title=""
|
||
function algorithm(n) {
|
||
var a = 1; // +1
|
||
a += 1; // +1
|
||
a *= 2; // +1
|
||
// 循环 n 次
|
||
for(let i = 0; i < n; i++){ // +1(每轮都执行 i ++)
|
||
console.log(0); // +1
|
||
}
|
||
}
|
||
```
|
||
|
||
=== "TypeScript"
|
||
|
||
```typescript title=""
|
||
function algorithm(n: number): void{
|
||
var a: number = 1; // +1
|
||
a += 1; // +1
|
||
a *= 2; // +1
|
||
// 循环 n 次
|
||
for(let i = 0; i < n; i++){ // +1(每轮都执行 i ++)
|
||
console.log(0); // +1
|
||
}
|
||
}
|
||
```
|
||
|
||
=== "C"
|
||
|
||
```c title=""
|
||
void algorithm(int n) {
|
||
int a = 1; // +1
|
||
a = a + 1; // +1
|
||
a = a * 2; // +1
|
||
// 循环 n 次
|
||
for (int i = 0; i < n; i++) { // +1(每轮都执行 i ++)
|
||
printf("%d", 0); // +1
|
||
}
|
||
}
|
||
```
|
||
|
||
=== "C#"
|
||
|
||
```csharp title=""
|
||
void algorithm(int n)
|
||
{
|
||
int a = 1; // +1
|
||
a = a + 1; // +1
|
||
a = a * 2; // +1
|
||
// 循环 n 次
|
||
for (int i = 0; i < n; i++) // +1(每轮都执行 i ++)
|
||
{
|
||
Console.WriteLine(0); // +1
|
||
}
|
||
}
|
||
```
|
||
|
||
=== "Swift"
|
||
|
||
```swift title=""
|
||
func algorithm(n: Int) {
|
||
var a = 1 // +1
|
||
a = a + 1 // +1
|
||
a = a * 2 // +1
|
||
// 循环 n 次
|
||
for _ in 0 ..< n { // +1
|
||
print(0) // +1
|
||
}
|
||
}
|
||
```
|
||
|
||
=== "Zig"
|
||
|
||
```zig title=""
|
||
|
||
```
|
||
|
||
=== "Dart"
|
||
|
||
```dart title=""
|
||
|
||
```
|
||
|
||
$T(n)$ 是一次函数,说明时间增长趋势是线性的,因此可以得出时间复杂度是线性阶。
|
||
|
||
我们将线性阶的时间复杂度记为 $O(n)$ ,这个数学符号称为「大 $O$ 记号 Big-$O$ Notation」,表示函数 $T(n)$ 的「渐近上界 Asymptotic Upper Bound」。
|
||
|
||
推算时间复杂度本质上是计算“操作数量函数 $T(n)$”的渐近上界。接下来,我们来看函数渐近上界的数学定义。
|
||
|
||
!!! abstract "函数渐近上界"
|
||
|
||
若存在正实数 $c$ 和实数 $n_0$ ,使得对于所有的 $n > n_0$ ,均有
|
||
$$
|
||
T(n) \leq c \cdot f(n)
|
||
$$
|
||
则可认为 $f(n)$ 给出了 $T(n)$ 的一个渐近上界,记为
|
||
$$
|
||
T(n) = O(f(n))
|
||
$$
|
||
|
||
![函数的渐近上界](time_complexity.assets/asymptotic_upper_bound.png)
|
||
|
||
从本质上讲,计算渐近上界就是寻找一个函数 $f(n)$ ,使得当 $n$ 趋向于无穷大时,$T(n)$ 和 $f(n)$ 处于相同的增长级别,仅相差一个常数项 $c$ 的倍数。
|
||
|
||
## 推算方法
|
||
|
||
渐近上界的数学味儿有点重,如果你感觉没有完全理解,也无需担心。因为在实际使用中,我们只需要掌握推算方法,数学意义可以逐渐领悟。
|
||
|
||
根据定义,确定 $f(n)$ 之后,我们便可得到时间复杂度 $O(f(n))$ 。那么如何确定渐近上界 $f(n)$ 呢?总体分为两步:首先统计操作数量,然后判断渐近上界。
|
||
|
||
### 1) 统计操作数量
|
||
|
||
针对代码,逐行从上到下计算即可。然而,由于上述 $c \cdot f(n)$ 中的常数项 $c$ 可以取任意大小,**因此操作数量 $T(n)$ 中的各种系数、常数项都可以被忽略**。根据此原则,可以总结出以下计数简化技巧:
|
||
|
||
1. **忽略与 $n$ 无关的操作**。因为它们都是 $T(n)$ 中的常数项,对时间复杂度不产生影响。
|
||
2. **省略所有系数**。例如,循环 $2n$ 次、$5n + 1$ 次等,都可以简化记为 $n$ 次,因为 $n$ 前面的系数对时间复杂度没有影响。
|
||
3. **循环嵌套时使用乘法**。总操作数量等于外层循环和内层循环操作数量之积,每一层循环依然可以分别套用上述 `1.` 和 `2.` 技巧。
|
||
|
||
以下示例展示了使用上述技巧前、后的统计结果。
|
||
|
||
$$
|
||
\begin{aligned}
|
||
T(n) & = 2n(n + 1) + (5n + 1) + 2 & \text{完整统计 (-.-|||)} \newline
|
||
& = 2n^2 + 7n + 3 \newline
|
||
T(n) & = n^2 + n & \text{偷懒统计 (o.O)}
|
||
\end{aligned}
|
||
$$
|
||
|
||
最终,两者都能推出相同的时间复杂度结果,即 $O(n^2)$ 。
|
||
|
||
=== "Java"
|
||
|
||
```java title=""
|
||
void algorithm(int n) {
|
||
int a = 1; // +0(技巧 1)
|
||
a = a + n; // +0(技巧 1)
|
||
// +n(技巧 2)
|
||
for (int i = 0; i < 5 * n + 1; i++) {
|
||
System.out.println(0);
|
||
}
|
||
// +n*n(技巧 3)
|
||
for (int i = 0; i < 2 * n; i++) {
|
||
for (int j = 0; j < n + 1; j++) {
|
||
System.out.println(0);
|
||
}
|
||
}
|
||
}
|
||
```
|
||
|
||
=== "C++"
|
||
|
||
```cpp title=""
|
||
void algorithm(int n) {
|
||
int a = 1; // +0(技巧 1)
|
||
a = a + n; // +0(技巧 1)
|
||
// +n(技巧 2)
|
||
for (int i = 0; i < 5 * n + 1; i++) {
|
||
cout << 0 << endl;
|
||
}
|
||
// +n*n(技巧 3)
|
||
for (int i = 0; i < 2 * n; i++) {
|
||
for (int j = 0; j < n + 1; j++) {
|
||
cout << 0 << endl;
|
||
}
|
||
}
|
||
}
|
||
```
|
||
|
||
=== "Python"
|
||
|
||
```python title=""
|
||
def algorithm(n: int) -> None:
|
||
a = 1 # +0(技巧 1)
|
||
a = a + n # +0(技巧 1)
|
||
# +n(技巧 2)
|
||
for i in range(5 * n + 1):
|
||
print(0)
|
||
# +n*n(技巧 3)
|
||
for i in range(2 * n):
|
||
for j in range(n + 1):
|
||
print(0)
|
||
```
|
||
|
||
=== "Go"
|
||
|
||
```go title=""
|
||
func algorithm(n int) {
|
||
a := 1 // +0(技巧 1)
|
||
a = a + n // +0(技巧 1)
|
||
// +n(技巧 2)
|
||
for i := 0; i < 5 * n + 1; i++ {
|
||
fmt.Println(0)
|
||
}
|
||
// +n*n(技巧 3)
|
||
for i := 0; i < 2 * n; i++ {
|
||
for j := 0; j < n + 1; j++ {
|
||
fmt.Println(0)
|
||
}
|
||
}
|
||
}
|
||
```
|
||
|
||
=== "JavaScript"
|
||
|
||
```javascript title=""
|
||
function algorithm(n) {
|
||
let a = 1; // +0(技巧 1)
|
||
a = a + n; // +0(技巧 1)
|
||
// +n(技巧 2)
|
||
for (let i = 0; i < 5 * n + 1; i++) {
|
||
console.log(0);
|
||
}
|
||
// +n*n(技巧 3)
|
||
for (let i = 0; i < 2 * n; i++) {
|
||
for (let j = 0; j < n + 1; j++) {
|
||
console.log(0);
|
||
}
|
||
}
|
||
}
|
||
```
|
||
|
||
=== "TypeScript"
|
||
|
||
```typescript title=""
|
||
function algorithm(n: number): void {
|
||
let a = 1; // +0(技巧 1)
|
||
a = a + n; // +0(技巧 1)
|
||
// +n(技巧 2)
|
||
for (let i = 0; i < 5 * n + 1; i++) {
|
||
console.log(0);
|
||
}
|
||
// +n*n(技巧 3)
|
||
for (let i = 0; i < 2 * n; i++) {
|
||
for (let j = 0; j < n + 1; j++) {
|
||
console.log(0);
|
||
}
|
||
}
|
||
}
|
||
```
|
||
|
||
=== "C"
|
||
|
||
```c title=""
|
||
void algorithm(int n) {
|
||
int a = 1; // +0(技巧 1)
|
||
a = a + n; // +0(技巧 1)
|
||
// +n(技巧 2)
|
||
for (int i = 0; i < 5 * n + 1; i++) {
|
||
printf("%d", 0);
|
||
}
|
||
// +n*n(技巧 3)
|
||
for (int i = 0; i < 2 * n; i++) {
|
||
for (int j = 0; j < n + 1; j++) {
|
||
printf("%d", 0);
|
||
}
|
||
}
|
||
}
|
||
```
|
||
|
||
=== "C#"
|
||
|
||
```csharp title=""
|
||
void algorithm(int n)
|
||
{
|
||
int a = 1; // +0(技巧 1)
|
||
a = a + n; // +0(技巧 1)
|
||
// +n(技巧 2)
|
||
for (int i = 0; i < 5 * n + 1; i++)
|
||
{
|
||
Console.WriteLine(0);
|
||
}
|
||
// +n*n(技巧 3)
|
||
for (int i = 0; i < 2 * n; i++)
|
||
{
|
||
for (int j = 0; j < n + 1; j++)
|
||
{
|
||
Console.WriteLine(0);
|
||
}
|
||
}
|
||
}
|
||
```
|
||
|
||
=== "Swift"
|
||
|
||
```swift title=""
|
||
func algorithm(n: Int) {
|
||
var a = 1 // +0(技巧 1)
|
||
a = a + n // +0(技巧 1)
|
||
// +n(技巧 2)
|
||
for _ in 0 ..< (5 * n + 1) {
|
||
print(0)
|
||
}
|
||
// +n*n(技巧 3)
|
||
for _ in 0 ..< (2 * n) {
|
||
for _ in 0 ..< (n + 1) {
|
||
print(0)
|
||
}
|
||
}
|
||
}
|
||
```
|
||
|
||
=== "Zig"
|
||
|
||
```zig title=""
|
||
|
||
```
|
||
|
||
=== "Dart"
|
||
|
||
```dart title=""
|
||
|
||
```
|
||
|
||
### 2) 判断渐近上界
|
||
|
||
**时间复杂度由多项式 $T(n)$ 中最高阶的项来决定**。这是因为在 $n$ 趋于无穷大时,最高阶的项将发挥主导作用,其他项的影响都可以被忽略。
|
||
|
||
以下表格展示了一些例子,其中一些夸张的值是为了强调“系数无法撼动阶数”这一结论。当 $n$ 趋于无穷大时,这些常数变得无足轻重。
|
||
|
||
<div class="center-table" markdown>
|
||
|
||
| 操作数量 $T(n)$ | 时间复杂度 $O(f(n))$ |
|
||
| ---------------------- | -------------------- |
|
||
| $100000$ | $O(1)$ |
|
||
| $3n + 2$ | $O(n)$ |
|
||
| $2n^2 + 3n + 2$ | $O(n^2)$ |
|
||
| $n^3 + 10000n^2$ | $O(n^3)$ |
|
||
| $2^n + 10000n^{10000}$ | $O(2^n)$ |
|
||
|
||
</div>
|
||
|
||
## 常见类型
|
||
|
||
设输入数据大小为 $n$ ,常见的时间复杂度类型包括(按照从低到高的顺序排列):
|
||
|
||
$$
|
||
\begin{aligned}
|
||
O(1) < O(\log n) < O(n) < O(n \log n) < O(n^2) < O(2^n) < O(n!) \newline
|
||
\text{常数阶} < \text{对数阶} < \text{线性阶} < \text{线性对数阶} < \text{平方阶} < \text{指数阶} < \text{阶乘阶}
|
||
\end{aligned}
|
||
$$
|
||
|
||
![时间复杂度的常见类型](time_complexity.assets/time_complexity_common_types.png)
|
||
|
||
!!! tip
|
||
|
||
部分示例代码需要一些预备知识,包括数组、递归算法等。如果遇到不理解的部分,请不要担心,可以在学习完后面章节后再回顾。现阶段,请先专注于理解时间复杂度的含义和推算方法。
|
||
|
||
### 常数阶 $O(1)$
|
||
|
||
常数阶的操作数量与输入数据大小 $n$ 无关,即不随着 $n$ 的变化而变化。
|
||
|
||
对于以下算法,尽管操作数量 `size` 可能很大,但由于其与数据大小 $n$ 无关,因此时间复杂度仍为 $O(1)$ 。
|
||
|
||
=== "Java"
|
||
|
||
```java title="time_complexity.java"
|
||
[class]{time_complexity}-[func]{constant}
|
||
```
|
||
|
||
=== "C++"
|
||
|
||
```cpp title="time_complexity.cpp"
|
||
[class]{}-[func]{constant}
|
||
```
|
||
|
||
=== "Python"
|
||
|
||
```python title="time_complexity.py"
|
||
[class]{}-[func]{constant}
|
||
```
|
||
|
||
=== "Go"
|
||
|
||
```go title="time_complexity.go"
|
||
[class]{}-[func]{constant}
|
||
```
|
||
|
||
=== "JavaScript"
|
||
|
||
```javascript title="time_complexity.js"
|
||
[class]{}-[func]{constant}
|
||
```
|
||
|
||
=== "TypeScript"
|
||
|
||
```typescript title="time_complexity.ts"
|
||
[class]{}-[func]{constant}
|
||
```
|
||
|
||
=== "C"
|
||
|
||
```c title="time_complexity.c"
|
||
[class]{}-[func]{constant}
|
||
```
|
||
|
||
=== "C#"
|
||
|
||
```csharp title="time_complexity.cs"
|
||
[class]{time_complexity}-[func]{constant}
|
||
```
|
||
|
||
=== "Swift"
|
||
|
||
```swift title="time_complexity.swift"
|
||
[class]{}-[func]{constant}
|
||
```
|
||
|
||
=== "Zig"
|
||
|
||
```zig title="time_complexity.zig"
|
||
[class]{}-[func]{constant}
|
||
```
|
||
|
||
=== "Dart"
|
||
|
||
```dart title="time_complexity.dart"
|
||
[class]{}-[func]{constant}
|
||
```
|
||
|
||
### 线性阶 $O(n)$
|
||
|
||
线性阶的操作数量相对于输入数据大小以线性级别增长。线性阶通常出现在单层循环中。
|
||
|
||
=== "Java"
|
||
|
||
```java title="time_complexity.java"
|
||
[class]{time_complexity}-[func]{linear}
|
||
```
|
||
|
||
=== "C++"
|
||
|
||
```cpp title="time_complexity.cpp"
|
||
[class]{}-[func]{linear}
|
||
```
|
||
|
||
=== "Python"
|
||
|
||
```python title="time_complexity.py"
|
||
[class]{}-[func]{linear}
|
||
```
|
||
|
||
=== "Go"
|
||
|
||
```go title="time_complexity.go"
|
||
[class]{}-[func]{linear}
|
||
```
|
||
|
||
=== "JavaScript"
|
||
|
||
```javascript title="time_complexity.js"
|
||
[class]{}-[func]{linear}
|
||
```
|
||
|
||
=== "TypeScript"
|
||
|
||
```typescript title="time_complexity.ts"
|
||
[class]{}-[func]{linear}
|
||
```
|
||
|
||
=== "C"
|
||
|
||
```c title="time_complexity.c"
|
||
[class]{}-[func]{linear}
|
||
```
|
||
|
||
=== "C#"
|
||
|
||
```csharp title="time_complexity.cs"
|
||
[class]{time_complexity}-[func]{linear}
|
||
```
|
||
|
||
=== "Swift"
|
||
|
||
```swift title="time_complexity.swift"
|
||
[class]{}-[func]{linear}
|
||
```
|
||
|
||
=== "Zig"
|
||
|
||
```zig title="time_complexity.zig"
|
||
[class]{}-[func]{linear}
|
||
```
|
||
|
||
=== "Dart"
|
||
|
||
```dart title="time_complexity.dart"
|
||
[class]{}-[func]{linear}
|
||
```
|
||
|
||
遍历数组和遍历链表等操作的时间复杂度均为 $O(n)$ ,其中 $n$ 为数组或链表的长度。
|
||
|
||
!!! question "如何确定输入数据大小 $n$ ?"
|
||
|
||
**数据大小 $n$ 需根据输入数据的类型来具体确定**。例如,在上述示例中,我们直接将 $n$ 视为输入数据大小;在下面遍历数组的示例中,数据大小 $n$ 为数组的长度。
|
||
|
||
=== "Java"
|
||
|
||
```java title="time_complexity.java"
|
||
[class]{time_complexity}-[func]{arrayTraversal}
|
||
```
|
||
|
||
=== "C++"
|
||
|
||
```cpp title="time_complexity.cpp"
|
||
[class]{}-[func]{arrayTraversal}
|
||
```
|
||
|
||
=== "Python"
|
||
|
||
```python title="time_complexity.py"
|
||
[class]{}-[func]{array_traversal}
|
||
```
|
||
|
||
=== "Go"
|
||
|
||
```go title="time_complexity.go"
|
||
[class]{}-[func]{arrayTraversal}
|
||
```
|
||
|
||
=== "JavaScript"
|
||
|
||
```javascript title="time_complexity.js"
|
||
[class]{}-[func]{arrayTraversal}
|
||
```
|
||
|
||
=== "TypeScript"
|
||
|
||
```typescript title="time_complexity.ts"
|
||
[class]{}-[func]{arrayTraversal}
|
||
```
|
||
|
||
=== "C"
|
||
|
||
```c title="time_complexity.c"
|
||
[class]{}-[func]{arrayTraversal}
|
||
```
|
||
|
||
=== "C#"
|
||
|
||
```csharp title="time_complexity.cs"
|
||
[class]{time_complexity}-[func]{arrayTraversal}
|
||
```
|
||
|
||
=== "Swift"
|
||
|
||
```swift title="time_complexity.swift"
|
||
[class]{}-[func]{arrayTraversal}
|
||
```
|
||
|
||
=== "Zig"
|
||
|
||
```zig title="time_complexity.zig"
|
||
[class]{}-[func]{arrayTraversal}
|
||
```
|
||
|
||
=== "Dart"
|
||
|
||
```dart title="time_complexity.dart"
|
||
[class]{}-[func]{arrayTraversal}
|
||
```
|
||
|
||
### 平方阶 $O(n^2)$
|
||
|
||
平方阶的操作数量相对于输入数据大小以平方级别增长。平方阶通常出现在嵌套循环中,外层循环和内层循环都为 $O(n)$ ,因此总体为 $O(n^2)$ 。
|
||
|
||
=== "Java"
|
||
|
||
```java title="time_complexity.java"
|
||
[class]{time_complexity}-[func]{quadratic}
|
||
```
|
||
|
||
=== "C++"
|
||
|
||
```cpp title="time_complexity.cpp"
|
||
[class]{}-[func]{quadratic}
|
||
```
|
||
|
||
=== "Python"
|
||
|
||
```python title="time_complexity.py"
|
||
[class]{}-[func]{quadratic}
|
||
```
|
||
|
||
=== "Go"
|
||
|
||
```go title="time_complexity.go"
|
||
[class]{}-[func]{quadratic}
|
||
```
|
||
|
||
=== "JavaScript"
|
||
|
||
```javascript title="time_complexity.js"
|
||
[class]{}-[func]{quadratic}
|
||
```
|
||
|
||
=== "TypeScript"
|
||
|
||
```typescript title="time_complexity.ts"
|
||
[class]{}-[func]{quadratic}
|
||
```
|
||
|
||
=== "C"
|
||
|
||
```c title="time_complexity.c"
|
||
[class]{}-[func]{quadratic}
|
||
```
|
||
|
||
=== "C#"
|
||
|
||
```csharp title="time_complexity.cs"
|
||
[class]{time_complexity}-[func]{quadratic}
|
||
```
|
||
|
||
=== "Swift"
|
||
|
||
```swift title="time_complexity.swift"
|
||
[class]{}-[func]{quadratic}
|
||
```
|
||
|
||
=== "Zig"
|
||
|
||
```zig title="time_complexity.zig"
|
||
[class]{}-[func]{quadratic}
|
||
```
|
||
|
||
=== "Dart"
|
||
|
||
```dart title="time_complexity.dart"
|
||
[class]{}-[func]{quadratic}
|
||
```
|
||
|
||
![常数阶、线性阶、平方阶的时间复杂度](time_complexity.assets/time_complexity_constant_linear_quadratic.png)
|
||
|
||
以「冒泡排序」为例,外层循环执行 $n - 1$ 次,内层循环执行 $n-1, n-2, \cdots, 2, 1$ 次,平均为 $\frac{n}{2}$ 次,因此时间复杂度为 $O(n^2)$ 。
|
||
|
||
$$
|
||
O((n - 1) \frac{n}{2}) = O(n^2)
|
||
$$
|
||
|
||
=== "Java"
|
||
|
||
```java title="time_complexity.java"
|
||
[class]{time_complexity}-[func]{bubbleSort}
|
||
```
|
||
|
||
=== "C++"
|
||
|
||
```cpp title="time_complexity.cpp"
|
||
[class]{}-[func]{bubbleSort}
|
||
```
|
||
|
||
=== "Python"
|
||
|
||
```python title="time_complexity.py"
|
||
[class]{}-[func]{bubble_sort}
|
||
```
|
||
|
||
=== "Go"
|
||
|
||
```go title="time_complexity.go"
|
||
[class]{}-[func]{bubbleSort}
|
||
```
|
||
|
||
=== "JavaScript"
|
||
|
||
```javascript title="time_complexity.js"
|
||
[class]{}-[func]{bubbleSort}
|
||
```
|
||
|
||
=== "TypeScript"
|
||
|
||
```typescript title="time_complexity.ts"
|
||
[class]{}-[func]{bubbleSort}
|
||
```
|
||
|
||
=== "C"
|
||
|
||
```c title="time_complexity.c"
|
||
[class]{}-[func]{bubbleSort}
|
||
```
|
||
|
||
=== "C#"
|
||
|
||
```csharp title="time_complexity.cs"
|
||
[class]{time_complexity}-[func]{bubbleSort}
|
||
```
|
||
|
||
=== "Swift"
|
||
|
||
```swift title="time_complexity.swift"
|
||
[class]{}-[func]{bubbleSort}
|
||
```
|
||
|
||
=== "Zig"
|
||
|
||
```zig title="time_complexity.zig"
|
||
[class]{}-[func]{bubbleSort}
|
||
```
|
||
|
||
=== "Dart"
|
||
|
||
```dart title="time_complexity.dart"
|
||
[class]{}-[func]{bubbleSort}
|
||
```
|
||
|
||
### 指数阶 $O(2^n)$
|
||
|
||
!!! note
|
||
|
||
生物学的“细胞分裂”是指数阶增长的典型例子:初始状态为 $1$ 个细胞,分裂一轮后变为 $2$ 个,分裂两轮后变为 $4$ 个,以此类推,分裂 $n$ 轮后有 $2^n$ 个细胞。
|
||
|
||
指数阶增长非常迅速,在实际应用中通常是不可接受的。若一个问题使用「暴力枚举」求解的时间复杂度为 $O(2^n)$ ,那么通常需要使用「动态规划」或「贪心算法」等方法来解决。
|
||
|
||
=== "Java"
|
||
|
||
```java title="time_complexity.java"
|
||
[class]{time_complexity}-[func]{exponential}
|
||
```
|
||
|
||
=== "C++"
|
||
|
||
```cpp title="time_complexity.cpp"
|
||
[class]{}-[func]{exponential}
|
||
```
|
||
|
||
=== "Python"
|
||
|
||
```python title="time_complexity.py"
|
||
[class]{}-[func]{exponential}
|
||
```
|
||
|
||
=== "Go"
|
||
|
||
```go title="time_complexity.go"
|
||
[class]{}-[func]{exponential}
|
||
```
|
||
|
||
=== "JavaScript"
|
||
|
||
```javascript title="time_complexity.js"
|
||
[class]{}-[func]{exponential}
|
||
```
|
||
|
||
=== "TypeScript"
|
||
|
||
```typescript title="time_complexity.ts"
|
||
[class]{}-[func]{exponential}
|
||
```
|
||
|
||
=== "C"
|
||
|
||
```c title="time_complexity.c"
|
||
[class]{}-[func]{exponential}
|
||
```
|
||
|
||
=== "C#"
|
||
|
||
```csharp title="time_complexity.cs"
|
||
[class]{time_complexity}-[func]{exponential}
|
||
```
|
||
|
||
=== "Swift"
|
||
|
||
```swift title="time_complexity.swift"
|
||
[class]{}-[func]{exponential}
|
||
```
|
||
|
||
=== "Zig"
|
||
|
||
```zig title="time_complexity.zig"
|
||
[class]{}-[func]{exponential}
|
||
```
|
||
|
||
=== "Dart"
|
||
|
||
```dart title="time_complexity.dart"
|
||
[class]{}-[func]{exponential}
|
||
```
|
||
|
||
![指数阶的时间复杂度](time_complexity.assets/time_complexity_exponential.png)
|
||
|
||
在实际算法中,指数阶常出现于递归函数。例如以下代码,不断地一分为二,经过 $n$ 次分裂后停止。
|
||
|
||
=== "Java"
|
||
|
||
```java title="time_complexity.java"
|
||
[class]{time_complexity}-[func]{expRecur}
|
||
```
|
||
|
||
=== "C++"
|
||
|
||
```cpp title="time_complexity.cpp"
|
||
[class]{}-[func]{expRecur}
|
||
```
|
||
|
||
=== "Python"
|
||
|
||
```python title="time_complexity.py"
|
||
[class]{}-[func]{exp_recur}
|
||
```
|
||
|
||
=== "Go"
|
||
|
||
```go title="time_complexity.go"
|
||
[class]{}-[func]{expRecur}
|
||
```
|
||
|
||
=== "JavaScript"
|
||
|
||
```javascript title="time_complexity.js"
|
||
[class]{}-[func]{expRecur}
|
||
```
|
||
|
||
=== "TypeScript"
|
||
|
||
```typescript title="time_complexity.ts"
|
||
[class]{}-[func]{expRecur}
|
||
```
|
||
|
||
=== "C"
|
||
|
||
```c title="time_complexity.c"
|
||
[class]{}-[func]{expRecur}
|
||
```
|
||
|
||
=== "C#"
|
||
|
||
```csharp title="time_complexity.cs"
|
||
[class]{time_complexity}-[func]{expRecur}
|
||
```
|
||
|
||
=== "Swift"
|
||
|
||
```swift title="time_complexity.swift"
|
||
[class]{}-[func]{expRecur}
|
||
```
|
||
|
||
=== "Zig"
|
||
|
||
```zig title="time_complexity.zig"
|
||
[class]{}-[func]{expRecur}
|
||
```
|
||
|
||
=== "Dart"
|
||
|
||
```dart title="time_complexity.dart"
|
||
[class]{}-[func]{expRecur}
|
||
```
|
||
|
||
### 对数阶 $O(\log n)$
|
||
|
||
与指数阶相反,对数阶反映了“每轮缩减到一半的情况”。对数阶仅次于常数阶,时间增长缓慢,是理想的时间复杂度。
|
||
|
||
对数阶常出现于「二分查找」和「分治算法」中,体现了“一分为多”和“化繁为简”的算法思想。
|
||
|
||
设输入数据大小为 $n$ ,由于每轮缩减到一半,因此循环次数是 $\log_2 n$ ,即 $2^n$ 的反函数。
|
||
|
||
=== "Java"
|
||
|
||
```java title="time_complexity.java"
|
||
[class]{time_complexity}-[func]{logarithmic}
|
||
```
|
||
|
||
=== "C++"
|
||
|
||
```cpp title="time_complexity.cpp"
|
||
[class]{}-[func]{logarithmic}
|
||
```
|
||
|
||
=== "Python"
|
||
|
||
```python title="time_complexity.py"
|
||
[class]{}-[func]{logarithmic}
|
||
```
|
||
|
||
=== "Go"
|
||
|
||
```go title="time_complexity.go"
|
||
[class]{}-[func]{logarithmic}
|
||
```
|
||
|
||
=== "JavaScript"
|
||
|
||
```javascript title="time_complexity.js"
|
||
[class]{}-[func]{logarithmic}
|
||
```
|
||
|
||
=== "TypeScript"
|
||
|
||
```typescript title="time_complexity.ts"
|
||
[class]{}-[func]{logarithmic}
|
||
```
|
||
|
||
=== "C"
|
||
|
||
```c title="time_complexity.c"
|
||
[class]{}-[func]{logarithmic}
|
||
```
|
||
|
||
=== "C#"
|
||
|
||
```csharp title="time_complexity.cs"
|
||
[class]{time_complexity}-[func]{logarithmic}
|
||
```
|
||
|
||
=== "Swift"
|
||
|
||
```swift title="time_complexity.swift"
|
||
[class]{}-[func]{logarithmic}
|
||
```
|
||
|
||
=== "Zig"
|
||
|
||
```zig title="time_complexity.zig"
|
||
[class]{}-[func]{logarithmic}
|
||
```
|
||
|
||
=== "Dart"
|
||
|
||
```dart title="time_complexity.dart"
|
||
[class]{}-[func]{logarithmic}
|
||
```
|
||
|
||
![对数阶的时间复杂度](time_complexity.assets/time_complexity_logarithmic.png)
|
||
|
||
与指数阶类似,对数阶也常出现于递归函数。以下代码形成了一个高度为 $\log_2 n$ 的递归树。
|
||
|
||
=== "Java"
|
||
|
||
```java title="time_complexity.java"
|
||
[class]{time_complexity}-[func]{logRecur}
|
||
```
|
||
|
||
=== "C++"
|
||
|
||
```cpp title="time_complexity.cpp"
|
||
[class]{}-[func]{logRecur}
|
||
```
|
||
|
||
=== "Python"
|
||
|
||
```python title="time_complexity.py"
|
||
[class]{}-[func]{log_recur}
|
||
```
|
||
|
||
=== "Go"
|
||
|
||
```go title="time_complexity.go"
|
||
[class]{}-[func]{logRecur}
|
||
```
|
||
|
||
=== "JavaScript"
|
||
|
||
```javascript title="time_complexity.js"
|
||
[class]{}-[func]{logRecur}
|
||
```
|
||
|
||
=== "TypeScript"
|
||
|
||
```typescript title="time_complexity.ts"
|
||
[class]{}-[func]{logRecur}
|
||
```
|
||
|
||
=== "C"
|
||
|
||
```c title="time_complexity.c"
|
||
[class]{}-[func]{logRecur}
|
||
```
|
||
|
||
=== "C#"
|
||
|
||
```csharp title="time_complexity.cs"
|
||
[class]{time_complexity}-[func]{logRecur}
|
||
```
|
||
|
||
=== "Swift"
|
||
|
||
```swift title="time_complexity.swift"
|
||
[class]{}-[func]{logRecur}
|
||
```
|
||
|
||
=== "Zig"
|
||
|
||
```zig title="time_complexity.zig"
|
||
[class]{}-[func]{logRecur}
|
||
```
|
||
|
||
=== "Dart"
|
||
|
||
```dart title="time_complexity.dart"
|
||
[class]{}-[func]{logRecur}
|
||
```
|
||
|
||
### 线性对数阶 $O(n \log n)$
|
||
|
||
线性对数阶常出现于嵌套循环中,两层循环的时间复杂度分别为 $O(\log n)$ 和 $O(n)$ 。
|
||
|
||
主流排序算法的时间复杂度通常为 $O(n \log n)$ ,例如快速排序、归并排序、堆排序等。
|
||
|
||
=== "Java"
|
||
|
||
```java title="time_complexity.java"
|
||
[class]{time_complexity}-[func]{linearLogRecur}
|
||
```
|
||
|
||
=== "C++"
|
||
|
||
```cpp title="time_complexity.cpp"
|
||
[class]{}-[func]{linearLogRecur}
|
||
```
|
||
|
||
=== "Python"
|
||
|
||
```python title="time_complexity.py"
|
||
[class]{}-[func]{linear_log_recur}
|
||
```
|
||
|
||
=== "Go"
|
||
|
||
```go title="time_complexity.go"
|
||
[class]{}-[func]{linearLogRecur}
|
||
```
|
||
|
||
=== "JavaScript"
|
||
|
||
```javascript title="time_complexity.js"
|
||
[class]{}-[func]{linearLogRecur}
|
||
```
|
||
|
||
=== "TypeScript"
|
||
|
||
```typescript title="time_complexity.ts"
|
||
[class]{}-[func]{linearLogRecur}
|
||
```
|
||
|
||
=== "C"
|
||
|
||
```c title="time_complexity.c"
|
||
[class]{}-[func]{linearLogRecur}
|
||
```
|
||
|
||
=== "C#"
|
||
|
||
```csharp title="time_complexity.cs"
|
||
[class]{time_complexity}-[func]{linearLogRecur}
|
||
```
|
||
|
||
=== "Swift"
|
||
|
||
```swift title="time_complexity.swift"
|
||
[class]{}-[func]{linearLogRecur}
|
||
```
|
||
|
||
=== "Zig"
|
||
|
||
```zig title="time_complexity.zig"
|
||
[class]{}-[func]{linearLogRecur}
|
||
```
|
||
|
||
=== "Dart"
|
||
|
||
```dart title="time_complexity.dart"
|
||
[class]{}-[func]{linearLogRecur}
|
||
```
|
||
|
||
![线性对数阶的时间复杂度](time_complexity.assets/time_complexity_logarithmic_linear.png)
|
||
|
||
### 阶乘阶 $O(n!)$
|
||
|
||
阶乘阶对应数学上的「全排列」问题。给定 $n$ 个互不重复的元素,求其所有可能的排列方案,方案数量为:
|
||
|
||
$$
|
||
n! = n \times (n - 1) \times (n - 2) \times \cdots \times 2 \times 1
|
||
$$
|
||
|
||
阶乘通常使用递归实现。例如以下代码,第一层分裂出 $n$ 个,第二层分裂出 $n - 1$ 个,以此类推,直至第 $n$ 层时终止分裂。
|
||
|
||
=== "Java"
|
||
|
||
```java title="time_complexity.java"
|
||
[class]{time_complexity}-[func]{factorialRecur}
|
||
```
|
||
|
||
=== "C++"
|
||
|
||
```cpp title="time_complexity.cpp"
|
||
[class]{}-[func]{factorialRecur}
|
||
```
|
||
|
||
=== "Python"
|
||
|
||
```python title="time_complexity.py"
|
||
[class]{}-[func]{factorial_recur}
|
||
```
|
||
|
||
=== "Go"
|
||
|
||
```go title="time_complexity.go"
|
||
[class]{}-[func]{factorialRecur}
|
||
```
|
||
|
||
=== "JavaScript"
|
||
|
||
```javascript title="time_complexity.js"
|
||
[class]{}-[func]{factorialRecur}
|
||
```
|
||
|
||
=== "TypeScript"
|
||
|
||
```typescript title="time_complexity.ts"
|
||
[class]{}-[func]{factorialRecur}
|
||
```
|
||
|
||
=== "C"
|
||
|
||
```c title="time_complexity.c"
|
||
[class]{}-[func]{factorialRecur}
|
||
```
|
||
|
||
=== "C#"
|
||
|
||
```csharp title="time_complexity.cs"
|
||
[class]{time_complexity}-[func]{factorialRecur}
|
||
```
|
||
|
||
=== "Swift"
|
||
|
||
```swift title="time_complexity.swift"
|
||
[class]{}-[func]{factorialRecur}
|
||
```
|
||
|
||
=== "Zig"
|
||
|
||
```zig title="time_complexity.zig"
|
||
[class]{}-[func]{factorialRecur}
|
||
```
|
||
|
||
=== "Dart"
|
||
|
||
```dart title="time_complexity.dart"
|
||
[class]{}-[func]{factorialRecur}
|
||
```
|
||
|
||
![阶乘阶的时间复杂度](time_complexity.assets/time_complexity_factorial.png)
|
||
|
||
## 最差、最佳、平均时间复杂度
|
||
|
||
**某些算法的时间复杂度不是固定的,而是与输入数据的分布有关**。例如,假设输入一个长度为 $n$ 的数组 `nums` ,其中 `nums` 由从 $1$ 至 $n$ 的数字组成,但元素顺序是随机打乱的;算法的任务是返回元素 $1$ 的索引。我们可以得出以下结论:
|
||
|
||
- 当 `nums = [?, ?, ..., 1]` ,即当末尾元素是 $1$ 时,需要完整遍历数组,此时达到 **最差时间复杂度 $O(n)$**;
|
||
- 当 `nums = [1, ?, ?, ...]` ,即当首个数字为 $1$ 时,无论数组多长都不需要继续遍历,此时达到 **最佳时间复杂度 $\Omega(1)$**;
|
||
|
||
“函数渐近上界”使用大 $O$ 记号表示,代表「最差时间复杂度」。相应地,“函数渐近下界”用 $\Omega$ 记号来表示,代表「最佳时间复杂度」。
|
||
|
||
=== "Java"
|
||
|
||
```java title="worst_best_time_complexity.java"
|
||
[class]{worst_best_time_complexity}-[func]{randomNumbers}
|
||
|
||
[class]{worst_best_time_complexity}-[func]{findOne}
|
||
```
|
||
|
||
=== "C++"
|
||
|
||
```cpp title="worst_best_time_complexity.cpp"
|
||
[class]{}-[func]{randomNumbers}
|
||
|
||
[class]{}-[func]{findOne}
|
||
```
|
||
|
||
=== "Python"
|
||
|
||
```python title="worst_best_time_complexity.py"
|
||
[class]{}-[func]{random_numbers}
|
||
|
||
[class]{}-[func]{find_one}
|
||
```
|
||
|
||
=== "Go"
|
||
|
||
```go title="worst_best_time_complexity.go"
|
||
[class]{}-[func]{randomNumbers}
|
||
|
||
[class]{}-[func]{findOne}
|
||
```
|
||
|
||
=== "JavaScript"
|
||
|
||
```javascript title="worst_best_time_complexity.js"
|
||
[class]{}-[func]{randomNumbers}
|
||
|
||
[class]{}-[func]{findOne}
|
||
```
|
||
|
||
=== "TypeScript"
|
||
|
||
```typescript title="worst_best_time_complexity.ts"
|
||
[class]{}-[func]{randomNumbers}
|
||
|
||
[class]{}-[func]{findOne}
|
||
```
|
||
|
||
=== "C"
|
||
|
||
```c title="worst_best_time_complexity.c"
|
||
[class]{}-[func]{randomNumbers}
|
||
|
||
[class]{}-[func]{findOne}
|
||
```
|
||
|
||
=== "C#"
|
||
|
||
```csharp title="worst_best_time_complexity.cs"
|
||
[class]{worst_best_time_complexity}-[func]{randomNumbers}
|
||
|
||
[class]{worst_best_time_complexity}-[func]{findOne}
|
||
```
|
||
|
||
=== "Swift"
|
||
|
||
```swift title="worst_best_time_complexity.swift"
|
||
[class]{}-[func]{randomNumbers}
|
||
|
||
[class]{}-[func]{findOne}
|
||
```
|
||
|
||
=== "Zig"
|
||
|
||
```zig title="worst_best_time_complexity.zig"
|
||
// 生成一个数组,元素为 { 1, 2, ..., n },顺序被打乱
|
||
pub fn randomNumbers(comptime n: usize) [n]i32 {
|
||
var nums: [n]i32 = undefined;
|
||
// 生成数组 nums = { 1, 2, 3, ..., n }
|
||
for (nums) |*num, i| {
|
||
num.* = @intCast(i32, i) + 1;
|
||
}
|
||
// 随机打乱数组元素
|
||
const rand = std.crypto.random;
|
||
rand.shuffle(i32, &nums);
|
||
return nums;
|
||
}
|
||
|
||
// 查找数组 nums 中数字 1 所在索引
|
||
pub fn findOne(nums: []i32) i32 {
|
||
for (nums) |num, i| {
|
||
// 当元素 1 在数组头部时,达到最佳时间复杂度 O(1)
|
||
// 当元素 1 在数组尾部时,达到最差时间复杂度 O(n)
|
||
if (num == 1) return @intCast(i32, i);
|
||
}
|
||
return -1;
|
||
}
|
||
```
|
||
|
||
=== "Dart"
|
||
|
||
```dart title="worst_best_time_complexity.dart"
|
||
[class]{}-[func]{randomNumbers}
|
||
|
||
[class]{}-[func]{findOne}
|
||
```
|
||
|
||
!!! tip
|
||
|
||
实际应用中我们很少使用「最佳时间复杂度」,因为通常只有在很小概率下才能达到,可能会带来一定的误导性。相反,「最差时间复杂度」更为实用,因为它给出了一个“效率安全值”,让我们可以放心地使用算法。
|
||
|
||
从上述示例可以看出,最差或最佳时间复杂度只出现在“特殊分布的数据”中,这些情况的出现概率可能很小,因此并不能最真实地反映算法运行效率。相较之下,**「平均时间复杂度」可以体现算法在随机输入数据下的运行效率**,用 $\Theta$ 记号来表示。
|
||
|
||
对于部分算法,我们可以简单地推算出随机数据分布下的平均情况。比如上述示例,由于输入数组是被打乱的,因此元素 $1$ 出现在任意索引的概率都是相等的,那么算法的平均循环次数则是数组长度的一半 $\frac{n}{2}$ ,平均时间复杂度为 $\Theta(\frac{n}{2}) = \Theta(n)$ 。
|
||
|
||
但在实际应用中,尤其是较为复杂的算法,计算平均时间复杂度比较困难,因为很难简便地分析出在数据分布下的整体数学期望。在这种情况下,我们通常使用最差时间复杂度作为算法效率的评判标准。
|
||
|
||
!!! question "为什么很少看到 $\Theta$ 符号?"
|
||
|
||
可能由于 $O$ 符号过于朗朗上口,我们常常使用它来表示「平均复杂度」,但从严格意义上看,这种做法并不规范。在本书和其他资料中,若遇到类似“平均时间复杂度 $O(n)$”的表述,请将其直接理解为 $\Theta(n)$ 。
|