Zeta (2) Function (1 Viewer)

U MAD BRO

Member
Joined
Jul 29, 2012
Messages
287
Gender
Undisclosed
HSC
N/A
What is the shortest way to prove the function because the one where you use sinx < x < tanx is really really long, elementary but tedious.

Is the proof using Fourier series the shortest one?


and why does = ?!?!?
 

Carrotsticks

Retired
Joined
Jun 29, 2009
Messages
9,494
Gender
Undisclosed
HSC
N/A
What is the shortest way to prove the function because the one where you use sinx < x < tanx is really really long, elementary but tedious

Is the proof using Fourier series the shortest one?


and why does = ?!?!?
I would say that the original proof (Euler) using the Taylor Series for the Sine function and equating coefficients is the shortest.

Also, you already have a copy of the proof for that indefinite integral. You use Integration by Parts and the integral reduces to an infinite series, namely zeta_2.
 

U MAD BRO

Member
Joined
Jul 29, 2012
Messages
287
Gender
Undisclosed
HSC
N/A
I would say that the original proof (Euler) using the Taylor Series for the Sine function and equating coefficients is the shortest.

Also, you already have a copy of the proof for that indefinite integral. You use Integration by Parts and the integral reduces to an infinite series, namely zeta_2.
Yes, the Euler proof looks both short and simple.

Do you use integration by parts to evaluate that integral? Is the working out simple for that integral?
 

Carrotsticks

Retired
Joined
Jun 29, 2009
Messages
9,494
Gender
Undisclosed
HSC
N/A
Oh found a little typo, but nothing that will entirely change the proof.
 

Carrotsticks

Retired
Joined
Jun 29, 2009
Messages
9,494
Gender
Undisclosed
HSC
N/A
Also in the proof above, I took it for granted that:



By dominance of functions, but it really should be properly proven (fairly straightforward).
 

Carrotsticks

Retired
Joined
Jun 29, 2009
Messages
9,494
Gender
Undisclosed
HSC
N/A
Since trials are over, I'll spend my time doing extra curricular stuff, they are so much fun =)

and I used this http://mathim.com/Mathstuff a guy taught me how to do do contour integration there ^_^
It's good to know this extra stuff because it's interesting, but it's probably best if you focus on the HSC for the moment. You have plenty of time for extra reading afterwards.
 

U MAD BRO

Member
Joined
Jul 29, 2012
Messages
287
Gender
Undisclosed
HSC
N/A
It's good to know this extra stuff because it's interesting, but it's probably best if you focus on the HSC for the moment. You have plenty of time for extra reading afterwards.
Yes, you're right, but I'll take a break for a week after trials doing extra curricular stuff, and going to the gym, should be a nice break from studying.
Then I'll get back to doing past papers all day =)
 

Carrotsticks

Retired
Joined
Jun 29, 2009
Messages
9,494
Gender
Undisclosed
HSC
N/A
Yep I would have done a proper proof using the Taylor Series too, but I would probably also specify for large x and for all n>0 since x^n can be more powerful than e^x within a fairly large domain, given n is large enough.
 

lolcakes52

Member
Joined
Oct 31, 2011
Messages
286
Gender
Undisclosed
HSC
2012
God I love taylor series. When I see a HSC question which uses the first few terms of sin(x) and Newton's method I'm like I SEE WHAT YOU DID THAR!
 
Joined
Sep 20, 2010
Messages
2,225
Gender
Undisclosed
HSC
2012
YES! And when it's like e^x < 1+x+x^2/2

But no one at my school has such an appreciation :(
 
Joined
Sep 20, 2010
Messages
2,225
Gender
Undisclosed
HSC
2012
Oh yes, sorry lol

Edit: Actually the question in my trial asked for fx= x-ln(1+x+x^2/2) and it was defined for x<0. It was a calculus inequality with stationary points etc and it ended up being e^x<1+x+x^2/2 for x<0. But for positive x, e^x > 1+x+x^2/2, yes.
 
Last edited:

Users Who Are Viewing This Thread (Users: 0, Guests: 1)

Top