[go: up one dir, main page]

login
Revision History for A334614 (Bold, blue-underlined text is an addition; faded, red-underlined text is a deletion.)

Showing entries 1-10 | older changes
a(n) = pi(prime(n) - n) + n, where pi is the prime counting function.
(history; published version)
#37 by N. J. A. Sloane at Sun Oct 04 23:33:29 EDT 2020
STATUS

proposed

approved

#36 by Omar E. Pol at Tue Sep 22 21:55:00 EDT 2020
STATUS

editing

proposed

#35 by Omar E. Pol at Tue Sep 22 21:54:07 EDT 2020
COMMENTS

It can be shown that a(n) > a(n-1) >= 1 and a(n) <= 2*n - 1 < 2*n (see proofs in LINKSthe Links section).

LINKS

Ya-Ping Lu, <a href="/A334614/a334614.pdf">Proofs of the two observations in COMMENTSthe Comments section</a>

STATUS

proposed

editing

Discussion
Tue Sep 22
21:55
Omar E. Pol: Minor edits.
#34 by Ya-Ping Lu at Tue Sep 22 21:51:14 EDT 2020
STATUS

editing

proposed

#33 by Ya-Ping Lu at Tue Sep 22 21:48:21 EDT 2020
COMMENTS

It can be shown that a(n) > a(n-1) >= 1 and a(n) <= 2*n - 1 < 2*n (see proofs in LINKS).

LINKS

Ya-Ping Lu, <a href="/A334614/a334614.pdf">Proofs of the two observations in COMMENTS</a>

STATUS

proposed

editing

Discussion
Tue Sep 22
21:50
Ya-Ping Lu: The proof of the two observations in COMMENTS is attached in LINKS.
#32 by Ya-Ping Lu at Mon Sep 14 04:11:16 EDT 2020
STATUS

editing

proposed

#31 by Ya-Ping Lu at Mon Sep 14 04:10:49 EDT 2020
COMMENTS

It can be noticed shown that this is an increasing sequence a(n) > a(n-1) >= 1 and the ratio of a(n) to <= 2*n is - 1 < 2. Short proofs of these observations are provided below*n.

Let p_n = prime(n).

i) a(n) > a(n-1) >= 1. a(n) - a(n-1) = (pi(p_n - n) + n) - (pi(p_(n-1) - (n-1)) + (n-1)) = pi(p_n - n) - pi(p_(n-1) - (n-1)) + 1. Since (p_n - n) - (p_(n-1) - (n-1)) = (p_n - p_(n-1) - 1 and p_n - p_(n-1) >= 1, we have pi(p_n -n) - pi(p_(n-1) - (n-1)) >= 0, or a(n) - a(n-1) >= 1. Thus a(n) > a(n - 1) > ... > a(1) = 1.

ii) a(n) < 2*n. It is shown that there is at least one prime in the range (k-pi(k), k) for any integer k > 2 (see comments in A332086). Replacing k by p_n, we know that there should be at least one prime in the range of (p_n - n, p_n), or pi(p_n) - pi(p_n - n) >= 1. Since pi(p_n - n) = n - (pi(p_n) - pi(p_n - n)) <= n - 1, we have a(n)/n = (pi(p_n - n) + n)/n = pi(p_n - n)/n + 1 <= (n - 1)/n + 1 = 2 - 1/n.

STATUS

proposed

editing

#30 by Ya-Ping Lu at Sat Sep 12 22:43:33 EDT 2020
STATUS

editing

proposed

Discussion
Sun Sep 13
03:02
Michel Marcus: Personally I am not convinced with these proofs, do we really need them ?
03:04
Michel Marcus: I would rather simply see 2 lines : a(n) > a(n-1) >= 1 and a(n) < 2*n, other opinions ?
04:29
Ya-Ping Lu: That's fine with me, Michel. The point I want to make here is they are not conjectures and can be proved. I can write up a full proof and attach it here if necessary.
#29 by Ya-Ping Lu at Sat Sep 12 22:38:03 EDT 2020
COMMENTS

ii) a(n)/n <= 2 - 1/*n. It is shown that there is at least one prime in the range (k-pi(k), k) for any integer k > 2 (see comments in A332086). Replacing k by p_n, we know that there should be at least one prime in the range of (p_n - n, p_n), or pi(p_n) - pi(p_n - n) >= 1. Since pi(p_n - n) = n - (pi(p_n) - pi(p_n - n)) <= n - 1, we have a(n)/n = (pi(p_n - n) + n)/n = pi(p_n - n)/n + 1 <= (n - 1)/n + 1 = 2 - 1/n.

STATUS

proposed

editing

Discussion
Sat Sep 12
22:43
Ya-Ping Lu: Yes, Michel. "a(n) < 2*n" is more straightforward.
#28 by Michel Marcus at Sat Sep 12 10:51:06 EDT 2020
STATUS

editing

proposed