s r 2 + 2 0 0 = r s 2 + 2 0 0 2
In the equation above, r and s are both rational numbers and their difference is an integer.
What is the number of solutions to the equation?
This section requires Javascript.
You are seeing this because something didn't load right. We suggest you, (a) try
refreshing the page, (b) enabling javascript if it is disabled on your browser and,
finally, (c)
loading the
non-javascript version of this page
. We're sorry about the hassle.
The equation becomes r 3 − s 3 = N ( N − 1 ) r s where N = 2 0 0 , and we want r = s + n for some integer n . Thus 3 s 2 n + 3 s n 2 + n 3 [ N ( N − 1 ) − 3 n ] s 2 + n [ N ( N − 1 ) − 3 n ] s − n 3 = N ( N − 1 ) ( s 2 + s n ) = 0 and so s = − 2 1 n ± 2 [ N ( N − 1 ) − 3 n ] n ( N ( N − 1 ) − 3 n ) ( N ( N − 1 ) + n ) Thus − N ( N − 1 ) ≤ n < 3 1 N ( N − 1 ) , and ( N ( N − 1 ) − 3 n ) ( N ( N − 1 ) + n ) must be a perfect square.
Simple calculations with N = 2 0 0 shows that there are only 5 values of n for which this is true: − 3 9 8 0 0 , − 3 9 0 0 0 , − 6 0 0 0 , 0 , 5 2 0 0 .
When n = − 3 9 8 0 0 , the only value of s is 1 9 9 0 0 . When n = 0 , the only value of s is 0 , and this is not a permitted solution of the original equation. The other three values of n each yield two values of s , and so there are a total of 7 solutions.
Could u explain me the calculations u did to get the 5 values for n?
Log in to reply
Set N= 200 into (N(N-1) - 3n) ( N(N-1) + n) = m^2 for some positive integer m .
Expand and simplify, then apply Simon's favorite factoring trick .
Log in to reply
@Pi Han Goh - I get m^2 + 3n^2 +79600n =1584040000 but how do i apply that trick to this type of equation
Log in to reply
@Maninder Dhanauta – Sorry, let me start over: expand my original equation gives
-3n^2 - 79600n + 1584040000 = m^2
since we are looking for integer roots, then the quadratic discriminant of the equation above (in n) is a perfect square:
79600^2 - 4(-3)(1584040000 - m^2) = c^2, for some integer c
Now can you solve it from here?
Log in to reply
@Pi Han Goh – You've got a Pell equation. That opens a whole new bag of difficulties!
Log in to reply
@Kazem Sepehrinia – No, we don't get Pell's equation.
We are looking to solve ( N ( N − 1 ) + n ) ( N ( N − 1 ) − 3 n ) = q 2 for integers n , q , where N = 2 0 0 . Expanding and collecting terms, this becomes [ 3 n + N ( N − 1 ) ] 2 + 3 q 2 = 4 N 2 ( N − 1 ) 2 From this it is clear that n and q must be of the same parity, so we can write 3 n + N ( N − 1 ) = 2 p − q and rewrite the condition in the form ( 2 p − q ) 2 + 3 q 2 p 2 − p q + q 2 ( p + q ω ) ( p + q ω 2 ) = 4 N 2 ( N − 1 ) 2 = N 2 ( N − 1 ) 2 = N 2 ( N − 1 ) 2 working in the Euclidean domain Z [ ω ] , where ω is the principal cube root of unity. Since 2 and 5 are both irreducible in this Euclidean domain, we deduce that N = 2 0 0 divides both p and q , and hence p = N P , q = N Q where ( P + Q ω ) ( P + Q ω 2 ) = 1 9 9 2 = ( 1 5 + 2 ω ) 2 ( 1 5 + 2 ω 2 ) 2 where the RHS is now a factorisation into irreducibles in Z [ ω ] ,
It now follows that P + Q ω must be equal to one of ( 1 5 + 2 ω ) 2 u , ( 1 5 + 2 ω 2 ) 2 u or 1 9 9 u , where u is one of the six units in Z [ ω ] . Running through these 1 8 cases, only a selection of them end up giving an integer value for n . Wtthout giving the details, we obtain − 3 9 0 0 0 , − 3 9 8 0 0 , − 6 0 0 0 , 0 , 5 2 0 0 as the only possible values of n .
The way @Pi Han Goh was going with this, he would obtain an equation with c 2 + 1 2 m 2 equal to a particular integer. This is basically the same as the one I have analysed, with an extra factor of 2 to worry about.
This approach gives us a strategy for dealing with the general case of other values of N . We have to factorise N ( N − 1 ) into irreducibles in Z [ ω ] ...
Log in to reply
@Mark Hennings – On second viewing, I discovered that my work (without using Mark's Euclidean domain thingy technique, because I haven't mastered it yet!!) still requires plenty of trial and error. But I did try to minimize the amount of trials needed. Here's my failed attempt:
From what I've written, we find to find non-negative integer solutions to c 2 + 1 2 m 2 = 2 5 3 4 4 6 4 0 0 0 0 .
Consider modulo 4, we can see that c 2 m o d 4 = 0 . Thus, let c = 2 c 1 and m = m 1 (for consistency sake). Then the equation simplifies to c 1 2 + 3 m 1 2 = 6 3 3 6 1 6 0 0 0 0 .
Consider modulo 5. The quadratic residues of 5 are 0,1 and 4. Since RHS is divisible by 5, then so does LHS. Thus c 1 and m 1 must be divisible by 5. Let c 1 = 5 c 2 , m 1 = 5 m 2 . The equation simplifies to c 2 2 + 3 m 2 2 = 2 5 3 4 4 6 4 0 0 .
Repeat the process again: Consider modulo 5. Since RHS is divisible by 5, then so does LHS. Thus c 2 and m 2 must be divisible by 5. Let c 2 = 5 c 3 , m 2 = 5 m 3 . The equation simplifies to c 3 2 + 3 m 3 2 = 1 0 1 3 7 8 5 6 = 3 1 8 4 2 .
Not sure how to finish this off (elegantly)... I was thinking of sum of 2 (or 4) squares theorem, but I'm pretty sure that will fail.
Problem Loading...
Note Loading...
Set Loading...
Let A = 2 0 0 2 − 2 0 0 . Thus r 3 − s 3 = A r s . Write r − s = n , where n ∈ Z . Then ( s + n ) 3 − s 3 = A ( s + n ) s , or ( A − 3 n ) s 2 + ( A n − 3 n 2 ) s − n 3 = 0 According to quadratic formula, we get s = − 2 n ± 2 ∣ n ∣ A − 3 n ( A + n ) ( A − 3 n ) Therefore, ( A + n ) ( A − 3 n ) = k 2 , where k is a non-negative integer. It follows that − A ≤ n ≤ ⌊ A / 3 ⌋ .
Note that A is divisible by 2 0 0 . In the equation ( A + n ) ( A − 3 n ) = k 2 look mod 5 , then − 3 n 2 ≡ k 2 ( m o d 5 ) . Hence n ≡ k ≡ 0 ( m o d 5 ) , since a perfect square mod 5 is congruent to 0 , 1 , 4 only. If you divide both sides of the equation by 2 5 and look mod 5 again, then it comes out that n ≡ k ≡ 0 ( m o d 2 5 ) .
In the equation ( A + n ) ( A − 3 n ) = k 2 look mod 8 , then − 3 n 2 ≡ k 2 ( m o d 8 ) . Hence n and k are both even, since a perfect square mod 8 is congruent to 0 , 1 , 4 only. If you divide both sides of the equation by 4 and look mod 8 again, then a similar argument gives n ≡ k ≡ 0 ( m o d 4 ) . Now write n = 4 n ′ and k = 4 k ′ and divide both sides of the equation by 1 6 to get ( A / 4 + n ′ ) ( A / 4 − 3 n ′ ) = k ′ 2 . Look mod 8 , then 4 + 4 n ′ + 5 n ′ 2 ≡ k ′ 2 ( m o d 8 ) . If n ′ is odd, then LHS of this congruence is 5 mod 8 , which is contradiction. Thus n ′ and k ′ are both even and finally n ≡ k ≡ 0 ( m o d 8 ) .
Thus in total we have that n ≡ k ≡ 0 ( m o d 2 0 0 ) . Write n = 2 0 0 N and k = 2 0 0 K to get ( 1 9 9 + N ) ( 1 9 9 − 3 N ) = K 2 For N = − 1 9 9 and N = 0 the LHS becomes a square. Then, for N ≡ 0 ( m o d 1 9 9 ) we have that d = g cd ( 1 9 9 + N , 1 9 9 − 3 N ) ∣ 3 ( 1 9 9 + N ) + 1 9 9 − 3 n = 4 × 1 9 9 Hence d = 1 , 2 , 4 .
If d = 1 , then 1 9 9 + N = a 2 and 1 9 9 − 3 N = b 2 , where a and b are co-prime odd positive integers. Hence 3 a 2 + b 2 = 7 9 6 . Its easy to see that a ≤ 1 5 . Checking a = 1 , 3 , 5 , . . . , 1 5 , reveals that a = 1 3 , 1 5 make 7 9 6 − 3 a 2 a perfect square. Thus we get two pairs: ( a , b ) = ( 1 3 , 1 7 ) , ( 1 5 , 1 1 ) that result in N = − 3 0 and N = 2 6 , respectively.
If d = 2 , then 1 9 9 + N = 2 a 2 and 1 9 9 − 3 N = 2 b 2 , where a and b are co-prime positive integers. It follows that 3 a 2 + b 2 = 3 9 8 . Its easy to see that a ≤ 1 1 . Check that none of a ≤ 1 1 make 3 9 8 − 3 a 2 a perfect square.
If d = 4 , then 1 9 9 + N = 4 a 2 and 1 9 9 − 3 N = 4 b 2 , where a and b are co-prime positive integers. It follows that 3 a 2 + b 2 = 1 9 9 . Its easy to see that a ≤ 7 . Checking a ≤ 7 , reveals that a = 1 makes 1 9 9 − 3 a 2 a perfect square. Thus we get one pair: ( a , b ) = ( 1 , 1 4 ) that results in N = − 1 9 8 .
Finally we get 5 solutions for n = 2 0 0 N , namely n = − 3 9 8 0 0 , − 3 9 0 0 0 , − 6 0 0 0 , 0 , 5 2 0 0 n = − 3 9 8 0 0 gives only s = 1 9 9 0 0 . n = 0 gives A = 0 , which is a contradiction! Other values of n give two different solutions for s . Thus the total number of solutions for the original equation is 7 .
Notice that if ( r , s ) is a solution, then ( − s , − r ) is a solution too. Thus I write four of the solutions to the original equation. ( r , s ) = ( s + n , s ) = ( − 1 9 9 0 0 , 1 9 9 0 0 ) , ( − 7 1 2 6 7 5 0 , 7 1 4 6 2 5 0 ) , ( − 1 7 1 2 0 0 0 , 1 7 9 0 0 0 0 ) , ( 1 1 6 7 6 0 0 , 1 1 1 0 4 0 0 )