1

What is wrong with this proof? According to the problem the claim of the proof is correct, but its not a sufficient proof. I'm a bit stumped on this one as to why it's not a valid proof, it seems so to me. I need to figure out where the issue is, I don't need to actually solve the proof itself but the fact that I can't find the issue worries me that maybe I don't even know its an issue and do it in my own proofs. Any help is appreciated. Thank you

Let $f(n) = n^3 +n$. We want to prove that $f(n) \in O(n^3)$. By the definition of Big-Oh, $f(n) \in O(n^3)$ iff $f(n) \leq c*n^3$ for some $c > 0$ and all $n\geq n_0$. Using the definition of $f(n)$, $n^3 + n \leq c*n^3$. Let $c=2$ and $n_0 = 1$. By substituting in, we have $n^3 + n \leq 2n^3$. So, subtracting an $n^3$ from both sides, we get $n\leq n^3$. Dividing by n for both sides, we get $1\leq n^2$ which is true for all $n\geq n_0$. So, $f(n) \in O(n^3)$

1 Answers1

1

For one, you suppose your conclusion from the start, which is usually not kosher in proof-writing, and it's generally hard to follow. That said, your logic in what you're trying to get to is okay - it's moreso the structure of your proof that's off to me than anything. Here's how I would rewrite your proof:

Notice, for all $n \ge 1$, we have $n \le n^3$. Therefore, $f(n) = n^3 + n \le n^3 + n^3 = 2n^3$. In our definition for big-O, this means our constant is $c=2$ and $n_0 = 1$. Thus by the previous calculation, $f \in \mathcal O(n^3)$

Alternatively, using a different definition of big-O, my preference would be that $f \in \mathcal O(g)$ iff $\limsup_{x \to a} |f(x)/g(x)|=1$. Then an alternative proof is

$$\limsup_{n \to \infty} \left| \frac{n^3 + n}{n^3} \right| = \limsup_{n \to \infty} \frac{1 + n^{-2}}{1} = 1$$ giving $n^3 + n \in \mathcal O (n^3)$. The steps between the two limits is that $n,n^3 \ge 0$ for $n \ge 0$, allowing us to remove the absolute value, and then we just divide by $n^3$ on top and bottom and calculate the obvious limit.

PrincessEev
  • 43,815
  • Thank you, I think it is the backwards reasoning/supposing conclusion from the start that is the issue here. And thank you for providing the re-written proof, that does look much nicer. – user65909 Apr 19 '20 at 02:11