If you find any mistakes, please make a comment! Thank you.

Solution to Understanding Analysis Exercise 2.6


Exercise 2.6.1

Let $\varepsilon > 0$ be arbitrary. Since $(x_n)$ converges to $x$, there exits $N\in\mathbf N$ such that \begin{equation}\label{eq2.6.1.1}|x_n-x|<\frac{\varepsilon}{2}\end{equation} for all $n\geqslant N$.
Whenever $n,m\geqslant N$, we have\begin{align*}|x_n-x_m|\leqslant &~|x_n-x|+|x-x_m|\\ \text{by }\eqref{eq2.6.1.1}\quad< &~ \frac{\varepsilon}{2}+\frac{\varepsilon}{2}=\varepsilon.\end{align*}Hence $(x_n)$ is a Cauchy sequence.


Exercise 2.6.2

(a) Consider the sequence $a_n$ \[a_n=(-1)^n\frac{1}{n}.\]It is clear convergent to $0$ and hence a Cauchy sequence.

(b) Impossible by Lemma 2.6.3.

(c) Impossible. If a monotone sequence has a Cauchy subsequence, then such a sequence has a convergent sequence. By Exercise 2.5.2 (d) this sequence converges.

(d) Consider the sequence $(a_n)$\[(1,0,2,0,3,0,4,0,\cdots).\]It is clear this sequence is unbounded. Moreover the subsequence $(a_{2n})$ is Cauchy.


Exercise 2.6.3

(a) Let $\varepsilon>0$ be arbitrary. Since $(x_n)$ and $(y_n)$ are Cauchy sequences, there exist $N_1,N_2\in\mathbf N$ such that\[|x_n-x_m|<\frac{\varepsilon}{2}\]for all $n,m> N_1$ and \[|y_n-y_m|<\frac{\varepsilon}{2}\]for all $n,m> N_2$.

Let $N=\max \{N_1,N_2\}$. For all $n,m > N$, then by triangle inequality we have \begin{align*}|(x_n+y_n)-(x_m+y_m)|=&~|(x_n-x_m)+(y_n-y_m)|\\ \leqslant &~ |x_n-x_m|+|y_n-y_m| \\ < &~ \frac{\varepsilon}{2}+\frac{\varepsilon}{2}=\varepsilon.\end{align*}Hence $(x_n+y_n)$ is a Cauchy sequence.

(b) Let $\varepsilon>0$ be arbitrary. Since $(x_n)$ and $(y_n)$ are Cauchy sequences, they are bounded by Lemma 2.6.3. Hence there exist $M>0$ such that $|x_n|< M$ and $|y_n|< M$ fro all $n\in\mathbf N$. Since $(x_n)$ and $(y_n)$ are Cauchy sequences, there exist $N_1,N_2\in\mathbf N$ such that\[|x_n-x_m|<\frac{\varepsilon}{2M}\]for all $n,m> N_1$ and \[|y_n-y_m|<\frac{\varepsilon}{2M}\]for all $n,m> N_2$.

Let $N=\max \{N_1,N_2\}$. For all $n,m > N$, then by triangle inequality we have \begin{align*}|x_ny_n-x_my_m|=&~|x_ny_n-x_my_n+x_my_n-x_my_m|\\ \leqslant &~ |x_ny_n-x_my_n|+|x_my_n-x_my_m| \\ =&~ |y_n||x_n-x_m|+|x_m||y_n-y_m|\\ < &~ M\cdot\frac{\varepsilon}{2M}+M\cdot\frac{\varepsilon}{2M}=\varepsilon.\end{align*}Hence $(x_ny_n)$ is a Cauchy sequence.


Exercise 2.6.4

(a) True. Since $(a_n)$ and $(b_n)$ are Cauchy, they are convergent. Hence by Algebraic Limit Theorem, the sequence $(a_n-b_n)$ is convergent. By Exercise 2.3.10 (b), the sequence $(|a_n-b_n|)$ is convergent as well. Therefore by Cauchy Criterion, $(|a_n-b_n|)$ is Cauchy.

(b) False. Counterexample: Let $a_n\equiv 1$ for all $n\in\mathbf N$.

(c) False. Counterexample: Let $a_n=(-1)^n\dfrac{1}{n+1}$, then it is easy to see that \[c_n=\begin{cases}0, &\text{if }n \text{ is even},\\ -1,&\text{if }n \text{ is odd}.\end{cases}\]Therefore $c_n$ is divergent and hence not Cauchy.


Exercise 2.6.5

(a) False. Take the harmonic series\[s_n=1+\frac{1}{2}+\frac{1}{3}+\cdots+\frac{1}{n}.\]Let $\varepsilon>0$ be arbitrary. Take $N> \dfrac{1}{\varepsilon}$. Then for all $n>N$, we have\[|s_{n+1}-s_n|=\frac{1}{n+1}<\frac{1}{N}<\varepsilon.\]Hence $(s_n)$ is pseudo-Cauchy and not bounded.

(b) True. It follows from the triangle inequality. Let $\varepsilon>0$ be arbitrary. Then there exist $N_1,N_2\in\mathbf N$ such that \[|x_{n+1}-x_n|<\frac{\varepsilon}{2}\]for all $n>N_1$ and \[|y_{n+1}-y_n|<\frac{\varepsilon}{2}\]for all $n>N_2$.

Let $N=\max\{N_1,N_2\}$. For all $n> N$, we have by triangle inequality that\begin{align*}|(x_{n+1}+y_{n+1})-(x_n+y_n)| \leqslant &~ |x_{n+1}-x_n|+|y_{n+2}-y_n|\\ <& ~\frac{\varepsilon}{2}+\frac{\varepsilon}{2}=\varepsilon.\end{align*}Hence $(x_n+y_n)$ is pseudo-Cauchy.


Exercise 2.6.6

Exercise 2.6.7

 

Linearity

This website is supposed to help you study Linear Algebras. Please only read these solutions after thinking about the problems carefully. Do not just copy these solutions.

Leave a Reply

Close Menu