Big 'Oh' notation.

Reb

Banned
Joined
Jan 25, 2012
Messages
3
If for two sequences \(\displaystyle f(n), g(n)\) with \(\displaystyle g(n)\geq 2\) we have \(\displaystyle f(n)=O(g(n))\),
then \(\displaystyle log(f(n))=O(log(g(n)))\).


Does this continue to hold if we allow \(\displaystyle g(n)<2\)?
 
If for two sequences \(\displaystyle f(n), g(n)\) with \(\displaystyle g(n)\geq 2\) we have \(\displaystyle f(n)=O(g(n))\),
then \(\displaystyle log(f(n))=O(log(g(n)))\).


Does this continue to hold if we allow \(\displaystyle g(n)<2\)?

If \(\displaystyle g(n)<2\), then \(\displaystyle g(n)=1\). But, \(\displaystyle \log(1) = 0\). I don't know much about Big O notation, but I would imagine that it doesn't work well as a function of zero.
 
I don't see anything requiring \(\displaystyle g(n)\in \mathbb{N}\) especially since \(\displaystyle \log g(n)\) wouldn't otherwise be.
 
I don't see anything requiring \(\displaystyle g(n)\in \mathbb{N}\) especially since \(\displaystyle \log g(n)\) wouldn't otherwise be.

As I said, I don't know much about big o notation. Typically, \(\displaystyle g(n)\) is used when the function is \(\displaystyle g:\mathbb{N} \to \)something, and \(\displaystyle g(x)\) is used when the input can be any real number. But, again, I don't know the notation at all.
 
Top