A sequence is defined recursively by
x n + 1 = 4 x n + 1 0 6 x n + 3 5
where x 0 = 1 0 . One can show that the sequence { x n } is given in closed form by
x n = c λ n − d a + b λ n
where a , b , c , and d are positive integers, and g cd ( a , b , c , d ) = 1 , and λ is a negative integer.
Find a + b + c + d + ∣ λ ∣
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.
An elegant solution. Just what we'd expect from @Mark Hennings
It's also interesting to find that the eigenvectors are related to the fixed points of the recursion (I.e. x=(6x+35)/(4x+10)).
Log in to reply
The matrix A can act on the space of rays through the origin in R 2 . The real number x corresponds to the ray { ( t t x ) : t ∈ R } , while the ray { ( t 0 ) : t ∈ R } corresponds to the point ∞ . Thus the transformation A on R 2 becomes the transformation A ^ from R ∪ { ∞ } to itself given by x ↦ 4 x + 1 0 6 x + 3 5 ∞ ↦ 2 3 Then x is a fixed point of A ^ precisely when ( 1 x ) is an eigenvector of A .
If we just want to find the values of a , b , c , d , and λ , without deriving them, we can use the actual values of x n and solve the simultaneous equations.
From x n = c λ n − d a + b λ n ⟹ n → ∞ lim x n = n → ∞ lim c λ n − d a + b λ n = n → ∞ lim c − λ n d λ n a + b = c b for ∣ λ ∣ > 1 . This implies that x n converges and let its limit be L . Then
x n + 1 ⟹ L 4 L 2 + 1 0 L 4 L 2 + 4 L − 3 5 ( 2 L − 5 ) ( 2 L + 7 ) ⟹ L c b = 4 x n + 1 0 6 x n + 3 5 = 4 L + 1 0 6 L + 3 5 = 6 L + 3 5 = 0 = 0 = 2 5 = 2 5 ⟹ c = 5 2 b when n → ∞ Since x n > 0 ∀ n ≥ 0 ⟹ L > 0
Now we have:
⎩ ⎪ ⎪ ⎪ ⎪ ⎪ ⎨ ⎪ ⎪ ⎪ ⎪ ⎪ ⎧ x 0 = 1 0 = c − d a + b x 1 = 1 0 1 9 = c λ − d a + b λ x 2 = 1 1 2 9 = c λ 2 − d a + b λ 2 ⟹ a + b = 1 0 c − 1 0 d ⟹ 1 0 a + 1 0 b λ = 1 9 c λ − 1 9 d ⟹ 1 1 a + 1 1 b λ 2 = 2 9 c λ 2 − 2 9 d . . . ( 1 ) . . . ( 2 ) . . . ( 3 )
{ ( 2 ) − 1 0 ( 1 ) : ( 3 ) − 1 1 ( 1 ) : 1 0 b ( λ − 1 ) = c ( 1 9 λ − 1 0 0 ) + 8 1 d 1 1 b ( λ 2 − 1 ) = c ( 2 9 λ 2 − 1 1 0 ) + 8 1 d . . . ( 4 ) . . . ( 5 )
⟹ ( 5 ) − ( 4 ) : b ( 1 1 λ 2 − 1 0 λ − 1 ) 5 5 λ 2 − 5 0 λ − 5 3 λ 2 + 1 2 λ − 1 5 λ 2 + 4 λ − 5 ( λ + 5 ) ( λ − 1 ) ⟹ λ = c ( 2 9 λ 2 − 1 9 λ − 1 0 ) = 5 8 λ 2 − 3 8 λ − 2 0 = 0 = 0 = 0 = − 5 Since c = 5 2 b Since λ < 0
From ( 4 ) : − 6 0 b = 5 2 b ( − 1 9 5 ) + 8 1 d ⟹ d = 9 2 b . From ( 1 ) : a + b = 4 b − 9 2 0 ⟹ a = 9 7 b . Then we have:
x n = 5 2 λ n − 9 2 9 7 + λ n = 1 8 λ n − 1 0 3 5 + 4 5 λ n
Therefore a + b + c + d + ∣ λ ∣ = 3 5 + 4 5 + 1 8 + 1 0 + 5 = 1 1 3 .
Let's consider the general bilinear recursion,
x n + 1 = c x n + d a x n + b
Pulling the a and c out, this becomes,
x n + 1 = c a ( x n + ( d / c ) ) ( x n + ( b / a ) )
which, after dividing the numerator by the denominator, becomes,
x n + 1 = c a ( 1 + x n + ( d / c ) ( b / a − d / c ) )
Adding ( d / c ) to both sides,
x n + 1 + ( d / c ) = c a + d + c a x n + ( d / c ) ( b / a − d / c )
Let y n = x n + ( d / c ) , then
y n + 1 = c a + d + c a y n ( b / a − d / c )
Now we'll make the magic substitution (Source: Introduction to Difference Equations, by Samuel Goldberg, 1958), for this problem, namely, we'll let y n = z n z n + 1 then, we have,
z n + 1 z n + 2 = c a + d + c a ( b / a − d / c ) z n + 1 z n
multiply through by z n + 1 , to get,
z n + 2 = c a + d z n + 1 + c a ( b / a − d / c ) z n
which is a standard second order linear difference equation, whose solution is readily available.
Let's substitute the given parameters of the original recursive equation, then,
z n + 2 = 4 6 + 1 0 z n + 1 + 4 6 ( 3 5 / 6 − 1 0 / 4 ) z n
which after simplifying becomes,
z n + 2 = 4 z n + 1 + 5 z n
or
z n + 2 − 4 z n + 1 − 5 z n = 0
The characteristic equation of this difference equation is,
m 2 − 4 m − 5 = 0
and this factors into ( m − 5 ) ( m + 1 ) = 0 . Thus, the solution of the difference equation is of the form,
z n = A ( − 1 ) n + B ( 5 ) n
Since x 0 = 1 0 , then y 0 = 1 0 + 2 5 = 2 2 5 , therefore, we can choose z 0 = 1 , then it follows that z 1 = 2 2 5
Now, we can construct two linear equations in the unknown coefficients A and B , namely,
1 = A + B , and
2 2 5 = − A + 5 B
Adding, gives us, 6 B = 2 2 7 , so that B = 1 2 2 7 , and therefore, A = 1 − 1 2 2 7 = − 4 5
Hence, z n = − 4 5 ( − 1 ) n + 1 2 2 7 ( 5 ) n , and it follows that y n is given by,
y n = − 1 5 ( − 1 ) n + 2 7 ( 5 ) n − 1 5 ( − 1 ) n + 1 + 2 7 ( 5 ) n + 1 = − 1 5 ( − 1 ) n + 2 7 ( 5 ) n 1 5 ( − 1 ) n + 1 3 5 ( 5 ) n = − 5 + 9 ( − 5 ) n 5 + 4 5 ( − 5 ) n
Therefore,
x n = y n − 2 5 = − 1 0 + 1 8 ( − 5 ) n 1 0 + 9 0 ( − 5 ) n + 2 5 − 4 5 ( − 5 ) n = 1 8 ( − 5 ) n − 1 0 3 5 + 4 5 ( − 5 ) n
which is in the form specified by the problem, hence λ = − 5 and a , b , c , d = 3 5 , 4 5 , 1 8 , 1 0 .
Therefore, the answer is 3 5 + 4 5 + 1 8 + 1 0 + 5 = 1 1 3
God, I love problems with magical solutions that I've never learned about.
Problem Loading...
Note Loading...
Set Loading...
Define sequence of vectors r n by the formulae r n + 1 = A r n = ( 6 4 3 5 1 0 ) r n n ≥ 0 r 0 = ( 1 1 0 ) Now the matrix A has eigensystem A ( 2 5 ) = 2 0 ( 2 5 ) A ( − 2 7 ) = − 4 ( − 2 7 ) Since ( 1 1 0 ) = 8 1 [ 9 ( 2 5 ) + 5 ( − 2 7 ) ] we deduce that r n = 8 1 [ 9 ( 2 5 ) 2 0 n + 5 ( − 2 7 ) ( − 4 ) n ] n ≥ 0 and x n is then the ratio of the first and second coefficients of r n , so that x n = 1 8 × 2 0 n − 1 0 × ( − 4 ) n 4 5 × 2 0 n + 3 5 × ( − 4 ) n = 1 8 ( − 5 ) n − 1 0 4 5 ( − 5 ) n + 3 5 n ≥ 0 so that a + b + c + d + ∣ λ ∣ = 4 5 + 3 5 + 1 8 + 1 0 + ∣ − 5 ∣ = 1 1 3 .