Timeline for Asymptotics of a Bernoulli-number-like function
Current License: CC BY-SA 2.5
6 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Jan 10, 2010 at 0:24 | comment | added | Michael Lugo | Reid, the comment box is probably too small to contain it, right? | |
| Jan 10, 2010 at 0:04 | comment | added | Reid Barton | I have a heuristic argument for the value of $C + \omega(n)$, which also predicts this kind of log-periodicity, but I won't have time to write it up tonight. | |
| Jan 9, 2010 at 23:47 | comment | added | Steve Huntsman | It would still be true that the double limit exists and equals unity, though. | |
| Jan 9, 2010 at 23:40 | comment | added | Michael Lugo | Your guess is certainly reasonable. I would have believed it myself if I hadn't taken the calculations out further. | |
| Jan 9, 2010 at 23:32 | comment | added | Steve Huntsman | Looks to me like for $k$ large $f(2,k) \approx f(n,k)$ for $n$ large. This is where my guess in the comments above comes from. | |
| Jan 9, 2010 at 23:18 | history | answered | Michael Lugo | CC BY-SA 2.5 |