算法导论3.2练习题

3.2-1

设$n_1\leqslant n_2;\\ \therefore f(n_1)\leqslant f(n_2),g(n_1)\leqslant g(n_2) \\ \therefore f(n_1)+g(n_1)\leqslant f(n_2)+g(n_2)\\ same;\\ \therefore f(g(n_1))\leqslant f(g(n_2))\\ same;\\ \because f(n)\ge 0 ,g(n)\ge 0 \\ \therefore f(n_1)\cdot g(n_1) \leqslant f(n_2)\cdot g(n_2)$

3.2-2

$\because a>0,b>0,c>0,n;\\ \therefore ac=ca\\ \therefore log_b a \cdot log_b c =log_b c \cdot log_b a \\ \therefore log_b a^{log_b c}=log_b c^{log_b a} \\ \therefore a^{log_bc}=c^{log_ba}$

3.2-3

$\forall c>0,\exists n_0 >0,\\ make \space \forall n \ge n_0 ,\\ \therefore \exists 0\le n!\le cn^n \\ \therefore n!=\omega(n^n)\\ \ \\ \forall c>0,\exists n_0 >0,\\ make \space \forall n \ge n_0 ,\\ \therefore \exists 0\le c2^n <n! \\ \therefore n!=\omega(2^n) \\ \ \\ \because n!=\sqrt{2\pi n}(\frac n e)^n(1+\Theta(\frac1 n))\\ \therefore lgn!=lg\sqrt{2\pi n}+lg((\frac1 en)^n)+lg(1+\Theta(\frac 1 n)) \\ \therefore lgn!=\Theta(lgn^{\frac 1 2})+\Theta(nlgn)+ \Theta(lgn^{-1}) 此时将其他值转成常量不再关注 \\ \therefore lgn!=\Theta(lgn)+\Theta(nlgn)-\Theta(lgn) 将指数的符号取到外面抵消 \\ \therefore lgn!=\Theta(nlgn)$

3.2-4

$\because 函数的多项式有界定义：\\ \exists k : f(n)=O(n^k); \\ make:f(n)=\left \lceil lgn \right \rceil ! \\ \therefore \left \lceil lgn \right \rceil ! = c\cdot n^k; \\ \therefore \left \lceil lg 2^n \right \rceil ! = c \cdot (2^n)^k; \\ \therefore n!=c\cdot 2^{n\cdot k} ;\\ \therefore lg(n!)=lg(c)+lg(n\cdot k) ; \\ make: 忽略常量,并将k作为新的c: n! \neq O(n); \\ \therefore \left \lceil lgn \right \rceil ! 多项式无界$

$same :f(n)=\left \lceil \lg\space \lg n \right \rceil !;\\ \therefore \left \lceil \lg \space \lg(2^{2^{n}}) \right \rceil ! =c \cdot 2^{2^{n}k};\\ \therefore \lg(n !) =\lg(2^n k) ; \\ \therefore n! \neq O(2^n) ; \\ \therefore \left \lceil \lg \space \lg n \right \rceil ! 多项式无界！$

理论上，如果f(n)单调递增,且增速不小于$log_n$， 则$f(n)!$ 多项式无界

3.2-5

$因为函数lg^*的增长极慢，可以将其作为常量看,\\ \therefore \lg n<n;\\ \therefore 后者大$

3.2-6

分别将$\phi =\frac {1+\sqrt5} 2 ,\hat{\phi}=\frac{1-\sqrt5}2$代入到方程：$x^2=x+1$可得解

3.2-7

$make :i=0,thus:F_0=0;\\ make:i=1 , thus :F_1=1;\\ make: i > 1 , F_{ i - 1 } = \frac{ \phi ^{ i - 1} - \hat {\phi}^{i-1} } {\sqrt 5};\\ F_ i = \frac{ \phi^i - \hat {\phi}^i} {\sqrt 5};\\ F_{ i + 1 } = \frac{ \phi ^{ i + 1} - \hat {\phi}^{i+1} } {\sqrt 5};\\ $

3.2-8

$\because k\ln k=\Theta(n) (k作为常量存在); \\ \therefore n_0\ge k \ln k;\\ \because k=\Theta(\frac n {\ln n}); \\ \therefore n_1 \ge k \ln n \ge k \ln k \ln k = k \ln k +k \ln \ln k \\ n_0\le n_1 旧的渐进函数包涵了新的渐进函数的边界 \\ \therefore k= O(\frac n {\ln n }) ,k=\Omega(\frac n {\ln n}=0);\\ \therefore k=\Theta(\frac n {\ln n})$