Skip to content

Commit

Permalink
更新 LaTex 公式
Browse files Browse the repository at this point in the history
  • Loading branch information
itcharge committed May 12, 2024
1 parent d8789df commit 6b64aee
Showing 1 changed file with 5 additions and 5 deletions.
Original file line number Diff line number Diff line change
Expand Up @@ -68,7 +68,7 @@ def divide_and_conquer(problems_n): # problems_n 为问题规模

一般来讲,分治算法将一个问题划分为 $a$ 个形式相同的子问题,每个子问题的规模为 $n/b$,则总的时间复杂度的递归表达式可以表示为:

$T(n) = \begin{cases} \begin{array} \ \Theta{(1)} & n = 1 \cr a \times T(n/b) + f(n) & n > 1 \end{array} \end{cases}$
$T(n) = \begin{cases} \Theta{(1)} & n = 1 \cr a \times T(n/b) + f(n) & n > 1 \end{cases}$

其中,每次分解时产生的子问题个数是 $a$ ,每个子问题的规模是原问题规模的 $1 / b$,分解和合并 $a$ 个子问题的时间复杂度是 $f(n)$。

Expand All @@ -82,15 +82,15 @@ $T(n) = \begin{cases} \begin{array} \ \Theta{(1)} & n = 1 \cr a \times T(n/b) +

我们得出归并排序算法的递归表达式如下:

$T(n) = \begin{cases} \begin{array} \ O{(1)} & n = 1 \cr 2 \times T(n/2) + O(n) & n > 1 \end{array} \end{cases}$
$T(n) = \begin{cases} O{(1)} & n = 1 \cr 2 \times T(n/2) + O(n) & n > 1 \end{cases}$

根据归并排序的递归表达式,当 $n > 1$ 时,可以递推求解:

$\begin{align} T(n) & = 2 \times T(n/2) + O(n) \cr & = 2 \times (2 \times T(n / 4) + O(n/2)) + O(n) \cr & = 4 \times T(n/4) + 2 \times O(n) \cr & = 8 \times T(n/8) + 3 \times O(n) \cr & = …… \cr & = 2^x \times T(n/2^x) + x \times O(n) \end{align}$
$$\begin{aligned} T(n) & = 2 \times T(n/2) + O(n) \cr & = 2 \times (2 \times T(n / 4) + O(n/2)) + O(n) \cr & = 4 \times T(n/4) + 2 \times O(n) \cr & = 8 \times T(n/8) + 3 \times O(n) \cr & = …… \cr & = 2^x \times T(n/2^x) + x \times O(n) \end{aligned}$$

递推最终规模为 $1$,令 $n = 2^x$,则 $x = \log_2n$,则:

$\begin{align} T(n) & = n \times T(1) + \log_2n \times O(n) \cr & = n + \log_2n \times O(n) \cr & = O(n \times \log_2n) \end{align}$
$$\begin{aligned} T(n) & = n \times T(1) + \log_2n \times O(n) \cr & = n + \log_2n \times O(n) \cr & = O(n \times \log_2n) \end{aligned}$$

则归并排序的时间复杂度为 $O(n \times \log_2n)$。

Expand All @@ -106,7 +106,7 @@ $\text{时间复杂度} = \text{叶子数} \times T(1) + \text{成本和} = 2^x

归并排序算法的递归表达式如下:

$T(n) = \begin{cases} \begin{array} \ O{(1)} & n = 1 \cr 2T(n/2) + O(n) & n > 1 \end{array} \end{cases}$
$T(n) = \begin{cases} O{(1)} & n = 1 \cr 2T(n/2) + O(n) & n > 1 \end{cases}$

其对应的递归树如下图所示。

Expand Down

0 comments on commit 6b64aee

Please sign in to comment.