Can I decrease the factorization time with this formula ?

[CRGreathouse]

Originally Posted by Godzilla

Right, and for that, we need to develop an appropriate program for the precision, of the digits

It's very easy to find floor(N1), etc. with PARI/GP. You can use localprec as sm88 suggests, or sqrtint as I prefer, or you could just set the precision manually and use floor and sqrt directly. I'll give an example of each:

\\ Set local precision, then compute normally. Uses 9 guard digits.
N1(N)=localprec(logint(N,10)\2 + 9); floor(2*sqrt(.333 * N))

\\ Only integer operations are used, so precision is not an issue.
N1(N)=sqrtint(333*N\250)

\\ Set precision like \p 100 before calling this version.
\\ If not enough precision is used, an error will be produced.
N1(N)=floor(2*sqrt(.333 * N))

The difficulty is factoring N given these four constants. As far as I can tell all you're proposing is trial division, and a relatively inefficient version of trial division at that.

[Godzilla]

Originally Posted by CRGreathouse

The difficulty is factoring N given these four constants. As far as I can tell all you're proposing is trial division, and a relatively inefficient version of trial division at that.

For this i have the solution :

this source code written in C language (Super_number) , modifying the code....

http://www.corradodamiano.it/supernu...upernumeri.htm


.

[CRGreathouse]

At a glance that seems like an inefficient, homegrown bignum library. (I had one of those in college.) But that doesn't really address the core problem. Your factorization algorithm is too slow even at 64 bits where you don't need any bignum library at all.

[ET_]

Originally Posted by Godzilla

For this i have the solution :

this source code written in C language (Super_number) , modifying the code....

http://www.corradodamiano.it/supernu...upernumeri.htm


.

Fossi in te proverei a studiarmi la libreria gmp (gmplib.org).
O il programma yafu, disponibile su sourceforge.

[Godzilla]

Originally Posted by ET_

Fossi in te proverei a studiarmi la libreria gmp (gmplib.org).
O il programma yafu, disponibile su sourceforge.

Ho chiesto sul sito iprogrammatori.it proprio oggi e appunto come dici tu bisogna saper programmare un programma ad hoc. Io sinceramente riuscirei solo ad implementare quello postato sul link e non con poche difficoltà.



---EDIT---

and trying with the Turing machine Z program ?? Link files : https://github.com/adamyedidia/parsimony , and Link Video Tutorial : https://www.youtube.com/watch?v=Rt_SfOkk0YM and Link Video Explanation : https://www.youtube.com/watch?v=QtO8EWwggx4


.

[ET_]

Originally Posted by Godzilla

Ho chiesto sul sito iprogrammatori.it proprio oggi e appunto come dici tu bisogna saper programmare un programma ad hoc. Io sinceramente riuscirei solo ad implementare quello postato sul link e non con poche difficoltà.



---EDIT---

and trying with the Turing machine Z program ?? Link files : https://github.com/adamyedidia/parsimony , and Link Video Tutorial : https://www.youtube.com/watch?v=Rt_SfOkk0YM and Link Video Explanation : https://www.youtube.com/watch?v=QtO8EWwggx4


.

Il problema, qui, e' la mancanza di specifiche tecniche sul tuo algoritmo: se riesci a descrivere l'algoritmo in funzione di cio' che fai (e del perche' funziona), scrivere un programma è facile.
Considera anche un ripasso della teoria della complessità: ti permetterà di comprendere meglio in che modo "contare" le operazioni necessarie al completamento di un task, e di come l'esecuzione delle operazioni sia legata al tempo da funzioni diverse (logaritmica, lineare, polinomiale o esponenziale).

The problem here is on the lack of technical specifications about your algorithm: once you describe the algorithm in function of what you do (and why the result is correct), programming it will be easy.
Also consider studying some complexity theory: it will explain you better how to count operations and how operations are related to time by a function that can be logarithmic, linear, polynomial or exponential.

[westicles]

Can you guys bring back Silverman? He was great on threads like this.

[LaurV]

Originally Posted by westicles

Can you guys bring back Silverman? He was great on threads like this.

My feeling too. I don't understand how so many intelligent people waste their time with such an idiot and or troll. At least don blazys was funny.

[kladner]

This thread most likely would not have made it past the first page, and that page would be blood-spattered.

[Godzilla]

Originally Posted by LaurV

My feeling too. I don't understand how so many intelligent people waste their time with such an idiot and or troll. At least don blazys was funny.

I wrote this song is Dark sol3 Feel it ,https://www.youtube.com/watch?v=dHTIX1CxWE4 and listen LaurV... Do you like ?