Let f(x) = 2x^3 + 3x log(x). Prove f(x) is O(x^3) by finding values for C and k that satisfy the definition of big-O. Def: |f(x)| <= C|g(x)| , for all x >= k
@shubhamsrg
ive got C=3 and k=1, but i think my "proof" gets flawed
Please show your working.
Try C = 5 instead :3
Sir first please please please show your working
2x^3 + 3xlog(x) < 2x^3 + 3x^2 < 2x^3 + 3x^3 for all x >=1
Let Cx^3 be a linear combination similar to that of f(x). Cx^3 = 2x^3 + Nx^3 ; now find a suitable N such that: Nx^3 > 3x log(x) Nx^3 > 3x log(x) ; divide by x Nx^2 > 3log(x) ; use log property, p log(q) = log(q^p) Nx^2 > log(x^3) ; base 2 each side ***************************************** 2^(Nx^2) > x^3 ; base 2 each side again ******************************************* 4N > 3, for all N > 3/4, let N = 1 Cx^3 = 2x^3 + x^3 = 3x^3, Let C = 3 ..................................... now to solve for x 3x^3 > 2x^3 + 3xlog(x) x^3 > 3xlog(x) x^2 > log(x^3) 2^(2x) > x^3 4x > 3x, for all x > 0, let k = 1 or more
trying to steer away from log(x) is O(x)
ok
I've actually never heard of this O stuff until now :D
it pretty vague to me too. Just havent had a class that covers it is all
Logs are base 2 in this i believe since its from a computer science course \[Nx^2~\gt~log_{2}(x^3)\] \[2^{Nx^2}~\gt~2^{log(x^3)}\] \[2^{Nx^2}~\gt~x^3\] \[\Large 2^{2^{Nx^2}}~\gt~2^{x^3}\] \[2Nx^2~\gt~3x\] \[2Nx~\gt~3\] i think thats the way my proof should go, i just assumed that extra 2 base multipled exponentts into 4Nx
N > 3/(2x) when x=1, N >= 2 would work fine .... and of course C and k need not be unique solutions, just valid solutions
let C = 2 + 2 = 4 4x^3 > 2x^3 + 3x log(x) 2x^3 > 3x log(x) 2x^2 > 3 log(x) 6x^2 > 3x 2x > 1, when x > 1/2, let k=1
<cringing in horror> O.O
ugh, im still having an issue trying to base out the logs ... \[\Large N^{p^q}\ne N^{pq}\] \[\Large (N^p)^q= N^{pq}\]
Why were you steering clear of log(x) = O(x) again?
im not sure if the question presupposes that knowledge
Well, maybe not that, but isn't log(x) < x for all x > 1 anyway?
putting aside all the O stuff... log(x) is simply less than x?
granted we could prove it \[x>\frac{L(x)}{L(2)}\] \[x~L(2)>L(x);~~\forall>1\]
lost an x lol
And if you could prove it, then, the part I posted would follow? 2x^3 + 3x log(x) < 2x^3 + 3x^2 < 2x^3 + 3x^3 for all x > 1 (or maybe 2, just to be safe)
yes, that would follow if it was be proved that log(x) is O(x).
Really? It has to be that? and not just log(x) < x ? T.T
it has to satisfy the definintion of big-O
Well, that depletes my ideas :D
but if it was proven before hand and assumed to be common knowledge at this point -- then its fine :) im just not sure how much common knowledge the person is that emailed this to me
Ask him? :D
eventually ....
have fun yall
Cheers :)
Join our real-time social learning platform and learn together with your friends!