mersenneforum.org Fun Sequence
 User Name Remember Me? Password
 Register FAQ Search Today's Posts Mark Forums Read

 2013-02-07, 10:36 #1 Sam Kennedy     Oct 2012 2·41 Posts Fun Sequence I've started searching this sequence for prime numbers: $((n+1)\mathrm{!}\ - \ n\mathrm{!})k + 1$ I like this sequence because N-1 is easy to factor so proving primality is straight forward. Are there any faster primality tests than rabin-miller for numbers like this? Has anyone searched a sequence similar to this?
 2013-02-07, 11:16 #2 axn     Jun 2003 34·67 Posts Your expression simplifies to k.n.n!+1. Why not ditch the extra n and search for k.n!+1? Anyways... Are you doing any sieving? I am not aware of anything faster than regular PRP tests (PFGW can do that). PFGW can also prove the primality with a -tm argument.
 2013-02-07, 11:35 #3 Sam Kennedy     Oct 2012 2·41 Posts I'm not sure how sieving applies to this, I know it's used in the other prime searching projects but I've never understood how it's actually used. The only thing I can think of is generating a bunch of the above factorials, trial dividing by small primes below a certain limit, then performing the primality tests on whichever don't divide.
 2013-02-07, 11:42 #4 axn     Jun 2003 10101001100112 Posts Sieving is merely a more efficient mechanism for doing trial factoring across a bunch of candidates, exploiting redundancies. And yes, sieving is applicable for your form, though I am not aware of any existing software that can do this (maybe MultiSieve?). But there are people in this forum who are capable of writing a sieve for this. Your search space is controlled by 2 parameters, k and n. What are the limits on these? Also, how do you plan to investigate them? Fix a k and search across different n's? Fix an n and search across different k's? Some other way? Depending upon your strategy, the sieving approach also would change. You could cut down your total search time upto 50% with a good siever. Last fiddled with by axn on 2013-02-07 at 11:47
 2013-02-07, 11:53 #5 Sam Kennedy     Oct 2012 2×41 Posts Depending on the complexity of the maths involved, I might be able to write my own sieve. The way I'm searching is fixing k and incrementing n. The limit for k is the same as the limit for an unsigned long, n is limited by the amount of memory needed to store the entire term. Are there any good references which could give me some ideas of coding a sieve?

 Similar Threads Thread Thread Starter Forum Replies Last Post kar_bon Aliquot Sequences 169 2022-11-06 18:03 devarajkandadai Miscellaneous Math 3 2020-12-01 22:08 sweety439 sweety439 17 2017-06-13 03:49 devarajkandadai Math 3 2007-03-20 19:43 Citrix Puzzles 5 2005-09-14 23:33

All times are UTC. The time now is 23:32.

Sat Dec 3 23:32:46 UTC 2022 up 107 days, 21:01, 0 users, load averages: 0.68, 0.84, 0.99

Copyright ©2000 - 2022, Jelsoft Enterprises Ltd.

This forum has received and complied with 0 (zero) government requests for information.

Permission is granted to copy, distribute and/or modify this document under the terms of the GNU Free Documentation License, Version 1.2 or any later version published by the Free Software Foundation.
A copy of the license is included in the FAQ.

≠ ± ∓ ÷ × · − √ ‰ ⊗ ⊕ ⊖ ⊘ ⊙ ≤ ≥ ≦ ≧ ≨ ≩ ≺ ≻ ≼ ≽ ⊏ ⊐ ⊑ ⊒ ² ³ °
∠ ∟ ° ≅ ~ ‖ ⟂ ⫛
≡ ≜ ≈ ∝ ∞ ≪ ≫ ⌊⌋ ⌈⌉ ∘ ∏ ∐ ∑ ∧ ∨ ∩ ∪ ⨀ ⊕ ⊗ 𝖕 𝖖 𝖗 ⊲ ⊳
∅ ∖ ∁ ↦ ↣ ∩ ∪ ⊆ ⊂ ⊄ ⊊ ⊇ ⊃ ⊅ ⊋ ⊖ ∈ ∉ ∋ ∌ ℕ ℤ ℚ ℝ ℂ ℵ ℶ ℷ ℸ 𝓟
¬ ∨ ∧ ⊕ → ← ⇒ ⇐ ⇔ ∀ ∃ ∄ ∴ ∵ ⊤ ⊥ ⊢ ⊨ ⫤ ⊣ … ⋯ ⋮ ⋰ ⋱
∫ ∬ ∭ ∮ ∯ ∰ ∇ ∆ δ ∂ ℱ ℒ ℓ
𝛢𝛼 𝛣𝛽 𝛤𝛾 𝛥𝛿 𝛦𝜀𝜖 𝛧𝜁 𝛨𝜂 𝛩𝜃𝜗 𝛪𝜄 𝛫𝜅 𝛬𝜆 𝛭𝜇 𝛮𝜈 𝛯𝜉 𝛰𝜊 𝛱𝜋 𝛲𝜌 𝛴𝜎𝜍 𝛵𝜏 𝛶𝜐 𝛷𝜙𝜑 𝛸𝜒 𝛹𝜓 𝛺𝜔