Re: A problem about convergence of series.
- From: quasi <quasi@xxxxxxxx>
- Date: Wed, 14 Nov 2007 08:34:18 -0500
On Wed, 14 Nov 2007 07:20:12 -0600, David C. Ullrich
<ullrich@xxxxxxxxxxxxxxxx> wrote:
On 13 Nov 2007 21:39:52 -0500, hrubin@xxxxxxxxxxxxxxxxxxxx (Herman
Rubin) wrote:
In article <1194986239.224691.238050@xxxxxxxxxxxxxxxxxxxxxxxxxxxx>,
<131208@xxxxxxxxx> wrote:
f is twice differentiable on [0,1] (that is, f'' exists).
f'' is continuous on [0,1].
Then prove that
sum_{n = from 1 to infinite} f(1/n) converges iff f(0)=f'(0)=0.
All one needs is that f and f' exist on [0,1] (even a little
less can do) and f''(0) exists.
So for example if f(x) = 1 for all x then the sum of f(1/n)
converges?
I read Herman Rubin's reply as claiming this:
Let f : [0,1] -> R. If f is differentiable on [0,1], and f''(0)
exists, then
sum_{n = from 1 to infinity f(1/n) converges iff f(0) = f'(0) = 0.
quasi
.
- Follow-Ups:
- Re: A problem about convergence of series.
- From: David C . Ullrich
- Re: A problem about convergence of series.
- From: tommy1729
- Re: A problem about convergence of series.
- References:
- Re: A problem about convergence of series.
- From: Herman Rubin
- Re: A problem about convergence of series.
- From: David C . Ullrich
- Re: A problem about convergence of series.
- Prev by Date: Re: A problem about convergence of series.
- Next by Date: Re: [Translation] 'framing' of an expression ?
- Previous by thread: Re: A problem about convergence of series.
- Next by thread: Re: A problem about convergence of series.
- Index(es):
Relevant Pages
|