Simultaneous equations 4 unknowns.

apple2357

Junior Member
Joined
Mar 9, 2018
Messages
229
I have playing with this pair of equations and been looking for positive integer solutions.

a+b = c*d
c+d = a*b

Reason being is that there is a pretty solution found by guessing, a=1, b=5 c= 2 d= 3

We get

1+5= 2*3
2+3 =1*5

Nice?

Does anyone have any ideas on how i could either set up a spreadsheet or some other approach to find other solutions?
 

JeffM

Elite Member
Joined
Sep 14, 2012
Messages
3,849
I have never studied Diophantine equations so I cannot say if there is a simple way to deal with this problem.

\(\displaystyle a, \ b, \ c, \ d \in \mathbb Z^+,\ a + b = cd \text { and } c + d = ab. \)

\(\displaystyle a + b = cd \implies b = cd - a.\)

\(\displaystyle \therefore c + d = ab = a(cd - a) = acd - a^2 \implies a^2 - acd + c + d = 0 \implies\)

\(\displaystyle a = \dfrac{cd \pm \sqrt{(cd)^2 - 4(1)(c + d)}}{2} = \dfrac{cd \pm \sqrt{c^2d^2 - 4c - 4d}}{2}.\)

That implies that \(\displaystyle c^2d^2 - 4c - 4d = e^2, \text { such that } \sqrt{e^2} \in \mathbb Z.\)

You found a solution at c = 2 and d = 3. So cd = 6, c^2d^2 - 4c - 4d = 36 - 8 - 12 = 16. And this in turn gives (6+4)/2 = 5 and (6-4)/2 = 1 as solutions. So yes, you can use a spread sheet to generate additional answers, but a looper program might be better. Denis is good at those looper programs. I did a quick and dirty, and it appears that you have found the only examples in the smallest 25 positive integers, but I did not test my program intensively.

EDIT: Because of symmetry, 1 and 5 are also solutions, but I did not count those as different from 2 and 3. Probably wrong of me. In that case, you get solutions that permute the integers involved.
 

apple2357

Junior Member
Joined
Mar 9, 2018
Messages
229
Yes, my initial approach was just picking some values and assigning a=1, b=2 and working my way through that until i found positive integer solutions for c and d.
 

Subhotosh Khan

Super Moderator
Staff member
Joined
Jun 18, 2007
Messages
18,820
Diophantine equations are like Fermat's theorem - looks simple but .......
 
Top