Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > sci.physics > #546048

Re: Primality sieve challenge!

Show all headers | View raw

On 17/01/2016 8:45 PM, jonas.thornvall@gmail.com wrote:
> Maybe we could have a challenge finding the base with best reducion
> upto 100 000000? One probably do not want to store more vectors than
> that in memory i guess
>
> Reducing composites in the integer field
> http://jt.node365.se/composite.html
>
>
> I've made my own kind of primality sieve that work by reducing lthe
> number field into composite "legs" and prime "egs". The legs that
> contain just composites in the numberfield will be thrown away
> sieved?
>
> I asked the people at sci.math if there was a known upper limit for
> the reduction using this type of counters in the integer field, they
> said no.
>
> I wonder if there is a name for this type of sieve/counter within
> information theory?
>
> My idea is to find a base with very high percentage of composite
> legs, remove all start numbers that produce only composites in their
> legs, and then store the other egs that contain prime in an array.
>
> That array will than serve the purpose modelling the new integer
> field, using the searched base with highest composite reduction.
>
>
> The script seaching the bases take a few seconds, so there could be
> alot of improvements. I have a feeling that using bignumb bases as
> sieves will be out of range for us mere mortals, but who knows maybe
> for computer theorists.
>

If you're going to attach a special meaning to words, such as legs and 
egs (you've put them into quotes, after all), you should explain what 
you mean by them.

Sylvia.

Back to sci.physics | Previous | Next — Previous in thread | Find similar | Unroll thread

Thread

Primality sieve challenge! jonas.thornvall@gmail.com - 2016-01-17 01:45 -0800
  Re: Primality sieve challenge! jonas.thornvall@gmail.com - 2016-01-17 02:20 -0800
  Re: Primality sieve challenge! Sylvia Else <sylvia@not.at.this.address> - 2016-01-17 23:35 +1100

csiph-web