Determining the asymptotic relationship between two logarithmic functions.

51 Views Asked by At

This question is around computer science, but too mathmatichal to ask anywhere else. The log is base 2.

These are the two given functions:

$g(n) = 2^{\lg^2n}$

$f(n) = \sum_{i=1}^{n^2}\lg\left(i\right)$

After simplifying the term $2^{\lg^2n}$ to be $n^{\lg (n)}$ , so I can say that: $g(n) = \theta(n^{\lg (n)})$, and due to $f(n) = \sum_{i=1}^{n^2}\lg\left(i\right)$ being a sum function I can infer it is "boiling down" to the sum of an arithmetic sequence, which means I can say:

$f(n) = \theta(n^2)$


I am being asked to determine what are the asymptotic relationships between g and f.

I can infer both functions are of the same order, which is $n$ raised to the power of a number.

So, let's say I want to show that $g(n) = \theta(f(n))$:


So I have to show that there are positive constants known as $c_1$, $c_2$, and $n_0$, so that for every $n\ge n_0$, the following is true:

$c_1f(n) \le g(n) \le c_2f(n)$

I'll try to prove it for the right side of the inequality. I heard enlarging the expression and finding constants is the way to prove it. So, can I just enlarge the original expression to something big enough that will be $\le$ to the original one for every $n\ge 1$ ?

And that means I found $c_2 = 100$, and $n_0 = 1$?

$n^{\lg (n)} \le 100n^2$


Same line of taught will be on this side, but the opposite:

$0.5n^2 \le n^{\lg (n)}$

just taking a function that grows slow enough with the same order of $n^2$.

So it means there are $c_1=0.5$, and $n_0 = 1$?


And assuming it is done correctly, I can also infer that $g(n) = O(f(n))$, because I have, supposedly, found a $c=100$, and $n_0 = 1$ positive constants so that $g(n) \le cf(n)$ is true.

And by the same logic, infer $g(n) = \Omega(f(n))$, because I have, supposedly, found a $c=0.5$, and $n_0 = 1$ positive constants so that $cf(n) \le g(n)$.

So that means the asymptotic relationships between $g$ and $f$ are $O$ (big O), $\Omega$, and $\theta$?