New sequence : parallel "prime"

[CRGreathouse]

You have to read the last posts.
The algorithm was yet sent.

Sorry, I must have missed it in the flurry of posts.

I'm sorry I did a miscalculation.
I have to review the algorithm but it is not wrong in theory.
I"m tired now.
I go to sleep.
Tomorrow I will correct it.
You can correct it if you undesrstood the principle.

I do not understand the principle, but I'll look over the algorithm when you correct it (tomorrow or whenever is convenient).

[CRGreathouse]

billymac, I don't entirely understand your program (at least not well enough to speed it up), but here's my version of your code:

Code: Select all

parallelPrime2(lim=25)={
   my(syze=100000,arr,a=1,lim2=997,minn,v=[1]);
   arr=List(vector(syze, i, i));

   for(ctr=1,lim,
      ptr=a;
      move=2;

      forprime(move=2,lim2,
         ptr += move;      \\eg 1+2, then 1+3, then 1+5,1+7,1+11,...
         if(ptr>syze, error(ptr" = ptr > syze = "syze,k5));
         arr[ptr]=0;      \\strike-out sieve elements
      );

      minn=a*10;    \\Feb 6 tweak
      for(l7=1,syze,
         elem=arr[l7];
         if(elem<minn && elem>a,minn=elem)  \\want lowest nonzero value NOT already setaside
      );  \\end FOR l7
      a=minn;
      \\write("c:\\Pari\\momo3.txt",minn," , ")
      v=concat(v,minn)

   );  \\end while
   v
};

This outputs a vector rather than dumping it to a file, which is more convenient for me. Of course most of my changes were just code organization -- making it easier for me to see what's going on. I don't know if this makes it easier for you or harder; feel free to take anything you like and leave the rest.

momo: Feel free to take whichever program you prefer -- whichever makes more sense to you -- if you want a copy to run on your computer. I don't think I improved the runtime on this version, so if billymac's style makes more sense to you, by all means use it rather than this.

[momo]

You have to read the last posts.
The algorithm was yet sent.

Sorry, I must have missed it in the flurry of posts.

I'm sorry I did a miscalculation.
I have to review the algorithm but it is not wrong in theory.
I"m tired now.
I go to sleep.
Tomorrow I will correct it.
You can correct it if you undesrstood the principle.

I do not understand the principle, but I'll look over the algorithm when you correct it (tomorrow or whenever is convenient).

Hi,

I did not sleep yesterday. In fact the problem is more harder than I first thought.
So my last algo is not working.
To determine if a number n is parallel prime will not be easy.
Have good week end.
I'm going to stop posting one or two days (Pari/gp is very interesting).

Thank you for all your comments.

[CRGreathouse]

I'm going to stop posting one or two days (Pari/gp is very interesting).

It is! Feel free to start a post (here or on Math Software) about using Pari, if you run into any issues.

[momo]

Hi,

For now I did not find the solution for the question : is n parallel prime or not?
Because I have the first few values U(n) :
1,2,5,8,9,12,20,24,27,30,31,42,48,60,61,72,76,93,94,100,118,122,133,156,163...

I had discovered some pattern some regularities but it is not helpful.
n is not parallel prime if :

n=U(k) +s*(s+1)/2 - 1

where U(k) is a parallel prime and s natural number

I need a bigger list (at least 1000).

Help!!!

[CRGreathouse]

That property follows directly from your definition.

[momo]

That property follows directly from your definition.

It is obvious that it is a consequence of my definition.
But I need more longer list to try to estimate s and U(k).
I did it for three to four digit numbers and it is not sure that for big numbers my idea will hold.
There are some bounds to compute.

[CRGreathouse]

Well, perhaps billymac will adapt his program for you.

[billymac00]

Well, perhaps billymac will adapt his program for you.

naah, I've used up time allotment on this, I am studying for my PE, must stay focused!

[momo]

Well, perhaps billymac will adapt his program for you.

naah, I've used up time allotment on this, I am studying for my PE, must stay focused!

Hi,

I know you are busy but in this case you have to replace in your program the sequence prime 2.3.5.7....p by 2,3,4,5,6,7,8,.....n.
So I do not think that it is going to take you more than one minute.
Thank you if you can send it to me by PM.
You can also use Rapidshare
http://rapidshare.de/
You can upload up to 100 mega

Thank you!

[CRGreathouse]

naah, I've used up time allotment on this, I am studying for my PE, must stay focused!

Professional Engineer? Good luck with that, we're all rooting for you.

[billymac00]

OK, DONE as per request. the rapid share says link was [url]http://rapidshare.de/files/49127070/momonp.txt.html/url]
delete when done. this one is 461 data points, I'll let it run longer soon to get you 1000.

[momo]

OK, DONE as per request. the rapid share says link was [url]http://rapidshare.de/files/49127070/momonp.txt.html/url]
delete when done. this one is 461 data points, I'll let it run longer soon to get you 1000.

BIG BIG THANX!!!

[momo]

This night I'm going to work on the file

[billymac00]

This night I'm going to work on the file

Ok, here's better (longer) file, use it instead please...think it has 1443 entries...
[url]http://rapidshare.de/files/49127804/momonp2.txt.html/url]