My Math Forum  

Go Back   My Math Forum > College Math Forum > Number Theory

Number Theory Number Theory Math Forum


Reply
 
LinkBack Thread Tools Display Modes
April 3rd, 2009, 08:18 PM   #1
Senior Member
 
Joined: Nov 2007

Posts: 633
Thanks: 0

Diophantine equation : factorial and powers

Can we express n! as sum of powered numbers?

n! = (a^b + c^d + e^f ....) - (a'^b' + c'^d' + e'^f' ....)

where a,a',b,b',c,c' ..... are integers > 1
I'm not looking for a trivial solution.
The number of the variable used has to be minimal of course.

If we can find a general solution then the factorization problem will be over.
momo is offline  
 
April 3rd, 2009, 10:04 PM   #2
Global Moderator
 
CRGreathouse's Avatar
 
Joined: Nov 2006
From: UTC -5

Posts: 16,046
Thanks: 938

Math Focus: Number theory, computational mathematics, combinatorics, FOM, symbolic logic, TCS, algorithms
Re: Diophantine equation : factorial and powers

Every number can be so represented, factorials are no different from the others. The number of terms will be at least two.

2! = 2^3 - 2^2
3! = 13^3 - 3^7 - 2^2
4! = 2^10 - 10^3
5! = 31^2 - 29^2
6! = 181^2 - 179^2
7! = 1261^2 - 1259^2
8! = 5042^2 - 5038^2
9! = 610^2 - 96^2 - 2^2
10! = 78^4 - 322^3 - 2^3
11! = 6802^2 - 2520^2 - 2^2
12! = 6850^2 + 756^3 - 46^2


The code's bad, but if anyone wants to glance at it for some reason here it is. You could use it as a starting place if you like, but starting over may be better.
Code:
pw=[];for(n=2,1e4,for(k=2,log(1e8)/log(n),pw=concat(pw,n^k)));pw=vecsort(pw,,8);
lf(t,k)=for(i=k+1,#pw,if(pw[i]-pw[i-k]==t,return([pw[i],pw[i-k]])));0
lookfor(t)=for(k=1,(t>>20)+50,l=lf(t,k);if(l!=0,return(l)))
nice(v)=my(a,b,c,d);if(v==0,print("?");return());b=ispower(v[1],,&a);d=ispower(v[2],,&c);print(a"^"b" - "c"^"d)
for(n=1,30,print1(n"! = ");nice(lookfor(n!)))

ff(t)=for(i=1,#pw-1,for(j=i+1,#pw,if(ispower(pw[j]-pw[i]-t),return([pw[j],pw[i]]))));
nice2(n)=my(a,b,c,d,e,f,v=ff(n!));b=ispower(v[1],,&a);d=ispower(v[2],,&c);f=ispower(v[1]-v[2]-n!,,&e);if(e>0,print(n"! = "a"^"b" - "e"^"f" - "c"^"d),print(n"! = "a"^"b" + "-e"^"f" - "c"^"d))
CRGreathouse is offline  
April 4th, 2009, 06:52 AM   #3
Senior Member
 
Joined: Nov 2007

Posts: 258
Thanks: 0

Re: Diophantine equation : factorial and powers

Quote:
Originally Posted by momo
Can we express n! as sum of powered numbers?

n! = (a^b + c^d + e^f ....) - (a'^b' + c'^d' + e'^f' ....)

where a,a',b,b',c,c' ..... are integers > 1
I'm not looking for a trivial solution.
The number of the variable used has to be minimal of course.

If we can find a general solution then the factorization problem will be over.
Every integer is the sum of four squares. That is nontrivial. Hence yes, you can always write



for some integers a,b,c,d. Or you can use 9 cubes, or 19 fourth powers.

Quote:
the factorization problem will be over
I am tired of hearing claims like this. You don't seem to be very knowledgeable in mathematics. Usually, when somebody claims to have a solution to an important problem, the solution is wrong.
brunojo is offline  
April 4th, 2009, 06:59 AM   #4
Senior Member
 
Joined: Nov 2007

Posts: 633
Thanks: 0

Re: Diophantine equation : factorial and powers

Quote:
Originally Posted by brunojoyal
Quote:
Originally Posted by momo
Can we express n! as sum of powered numbers?

n! = (a^b + c^d + e^f ....) - (a'^b' + c'^d' + e'^f' ....)

where a,a',b,b',c,c' ..... are integers > 1
I'm not looking for a trivial solution.
The number of the variable used has to be minimal of course.

If we can find a general solution then the factorization problem will be over.
Every integer is the sum of four squares. That is nontrivial. Hence yes, you can always write



for some integers a,b,c,d. Or you can use 9 cubes, or 19 fourth powers.

Quote:
the factorization problem will be over
I am tired of hearing claims like this. You don't seem to be very knowledgeable in mathematics. Usually, when somebody claims to have a solution to an important problem, the solution is wrong.
Can you write this FACTORIAL number

16452142516874952231465221452111233522165123415213 65412987411256421236898741554125511122556345413661 123635645198744512456314145514556541!

as sum of 4 squares?

I know we can do in theory but how to solve it practically?
momo is offline  
April 4th, 2009, 07:04 AM   #5
Senior Member
 
Joined: Nov 2007

Posts: 633
Thanks: 0

Re: Diophantine equation : factorial and powers

What I'm talking about is very simple

If we can express any factorial of one big big number (100000 digits or more) as :

a^b + c^d (for example) then you can factorize any number < to this number.

That's it.

I'm not claiming nothing.

I close this post anyway.
momo is offline  
April 4th, 2009, 07:13 AM   #6
Global Moderator
 
CRGreathouse's Avatar
 
Joined: Nov 2006
From: UTC -5

Posts: 16,046
Thanks: 938

Math Focus: Number theory, computational mathematics, combinatorics, FOM, symbolic logic, TCS, algorithms
Re: Diophantine equation : factorial and powers

Quote:
Originally Posted by brunojoyal
Every integer is the sum of four squares. That is nontrivial. Hence yes, you can always write



for some integers a,b,c,d. Or you can use 9 cubes, or 19 fourth powers.
True. I'm reading Nathanson's book on that right now, oddly enough.

But it's actually not quite enough here, since momo requires the base to be greater than 1. 0 isn't a problem, but 1 is. I suppose we can always add/subtract 9-8, though. So I guess you're right after all.
CRGreathouse is offline  
April 4th, 2009, 07:24 AM   #7
Global Moderator
 
CRGreathouse's Avatar
 
Joined: Nov 2006
From: UTC -5

Posts: 16,046
Thanks: 938

Math Focus: Number theory, computational mathematics, combinatorics, FOM, symbolic logic, TCS, algorithms
Re: Diophantine equation : factorial and powers

Quote:
Originally Posted by momo
Can you write this FACTORIAL number

16452142516874952231465221452111233522165123415213 65412987411256421236898741554125511122556345413661 123635645198744512456314145514556541!

as sum of 4 squares?

I know we can do in theory but how to solve it practically?
That number is too large to work with, having
22174514878431004692219364866552514726623549098554 39563429370449573913993660062043775137385374329379 79812945000881840846773887256567649950
digits. But there are efficient methods to decompose numbers into the sum of four squares; see for example
http://www.alpertron.com.ar/4SQUARES.HTM
CRGreathouse is offline  
May 8th, 2009, 12:00 PM   #8
Newbie
 
Joined: May 2009

Posts: 2
Thanks: 0

Re: Diophantine equation : factorial and powers

Quote:
Originally Posted by momo
Can we express n! as sum of powered numbers?

n! = (a^b + c^d + e^f ....) - (a'^b' + c'^d' + e'^f' ....)

where a,a',b,b',c,c' ..... are integers > 1
I'm not looking for a trivial solution.
The number of the variable used has to be minimal of course.

If we can find a general solution then the factorization problem will be over.
If n>=4, you can write n!=M*N where M and N are even numbers.

Let U = (M+N)/2 and V = (M-N)/2

You get: n!=M*N=(U+V)(U-V)=U^2-V^2

Example for n=7: 5040 = (2*3*4)*(5*6*7) = 24*210 = (117-93)*(117+93) = 117^2-93^2

Best regards,

Darío Alpern
Buenos Aires - Argentina
alpertron is offline  
May 8th, 2009, 12:21 PM   #9
Senior Member
 
Joined: Nov 2007

Posts: 633
Thanks: 0

Re: Diophantine equation : factorial and powers

Quote:
Originally Posted by alpertron
Quote:
Originally Posted by momo
Can we express n! as sum of powered numbers?

n! = (a^b + c^d + e^f ....) - (a'^b' + c'^d' + e'^f' ....)

where a,a',b,b',c,c' ..... are integers > 1
I'm not looking for a trivial solution.
The number of the variable used has to be minimal of course.

If we can find a general solution then the factorization problem will be over.
If n>=4, you can write n!=M*N where M and N are even numbers.

Let U = (M+N)/2 and V = (M-N)/2

You get: n!=M*N=(U+V)(U-V)=U^2-V^2

Example for n=7: 5040 = (2*3*4)*(5*6*7) = 24*210 = (117-93)*(117+93) = 117^2-93^2

Best regards,

Darío Alpern
Buenos Aires - Argentina
Thank you for your idea.
Wilson theorem state that if p is prime then (p-1!) mod p = p-1
My idea is :
If we can express (p-1!) as x^2 - y^2 we have to find x^2 or y^2 such as x^2 = z^(p-1) or y=t^(p-1) so we will have to solve only one of the two because z^(p-1) or t^(p-1) mod p = 1 (if p prime)
We have to write a general formula such as we can be sure to isolate z^(p-1) or t^(p-1)
By expanding (p-1)! we can find such form
(p-1)! = z^(p-1) - x^2
or
(p-1)! = y^2 - t^(p-1)

For some factorials it is possible to build such form.

Thank you for your idea.
momo is offline  
Reply

  My Math Forum > College Math Forum > Number Theory

Tags
diophantine, equation, factorial, powers



Search tags for this page
Click on a term to search for related topics.
Thread Tools
Display Modes


Similar Threads
Thread Thread Starter Forum Replies Last Post
An equation with consecutive powers Dacu Number Theory 3 May 31st, 2013 09:06 PM
An equation with the factorial Dacu Number Theory 5 May 25th, 2013 05:54 AM
diophantine equation islam Number Theory 4 January 11th, 2012 03:28 AM
equation with powers damgam Algebra 1 December 19th, 2008 03:17 PM
An equation with the factorial Dacu Algebra 2 December 31st, 1969 04:00 PM





Copyright © 2019 My Math Forum. All rights reserved.