See Original Golden Iterations I
and Original Golden Iterations II (updated)
You can find the value of ϕ = 2 1 + 5 by iteration. Sometimes, we want a faster algorithm.
In fact, we just need a way to find 5 !
Start with y 1 = 2 and y 2 = 2 1 ( y 1 + y 1 5 ) = 2 . 2 5 and for the ( n + 1 ) th iteration:
y n + 1 = 2 1 ( y n + y n 5 ) x n = 2 1 + y n
For large n , we can express the error term as ∣ e n ∣ = ∣ x n − ϕ ∣ . It can be shown that n → ∞ lim ∣ e n ∣ c ∣ e n + 1 ∣ = b a , where a , b are integers that do not have common divisors. Find a + b + c .
Note: A few terms in the iteration seeded with y 1 = 2 or x 1 = 1 . 5 :
x 1 = 1 . 5
x 2 = 1 . 6 2 5 0 0 0 0 0 0 0 0 0 0 0 0
x 3 = 1 . 6 1 8 0 5 5 5 5 5 5 5 5 5 5 6
x 4 = 1 . 6 1 8 0 3 3 9 8 8 9 5 7 9 0 2
x 5 = 1 . 6 1 8 0 3 3 9 8 8 7 4 9 8 9 5
...
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.
Great example!
e n = x n − φ = 2 1 ( y n − 5 )
e n + 1 = 2 1 ( y n + 1 − 5 ) = 2 1 ( 2 1 ( y n + y n 5 ) − 5 )
So e n c e n + 1 = ( 2 1 ( y n − 5 ) ) c 2 1 ( 2 1 ( y n + y n 5 ) − 5 )
= 2 c − 2 ( y n − 5 ) c y n + y n 5 − 2 5
If we set f n = y n − 5 , this expression can be written as
= 2 c − 2 f n c f n + f n + 5 5 − 5
= 2 c − 2 f n c ( f n + 5 ) f n 2
lim n → ∞ e n c e n + 1 = lim f → 0 2 c − 2 f n c ( f n + 5 ) f n 2 = lim f → 0 2 c − 2 f c 5 f 2
This limit has a nonzero value when c = 2 , and then evaluates to 5 1 .
a = 1 , b = 5 , c = 2 , a + b + c = 8
There are many ways to solve this. Post your solutions!
The following is just one example:
Let f ( y ) = 2 1 ( y + 5 / y ) . The algorithm is a fixed point iteration with y n + 1 = f ( y n ) and y ∗ = 5 is the fixed point.
We also notice that d y d f at y ∗ is equal to 2 1 ( 1 − ( y ∗ ) 2 5 ) = 0 . Thus, for y n 's near y ∗ we can expand f ( y n ) as f ( y ∗ ) + 2 1 d y 2 d 3 f ( y ∗ ) ( y n − y ∗ ) 2 .
Thus,
f ( y n ) − f ( y ∗ ) = 2 1 d y 2 d 2 f ( y ∗ ) ( y n − y ∗ ) 2
2 1 [ ( f ( y n ) − f ( y ∗ ) ] = d y 2 d 2 f ( y ∗ ) [ 2 1 ( y n − y ∗ ) ] 2
We also note that
e n = x n − ϕ = 2 1 ( y n − y ∗ )
and
e n + 1 = x n + 1 − ϕ = 2 1 ( y n + 1 − y ∗ ) = 2 1 ( f ( y n ) − f ( y ∗ ) ) .
Thus, ∣ e n + 1 ∣ = ∣ d y 2 d 2 f ( y ∗ ) ∣ ∣ e n ∣ 2 , and gives c = 2.
∣ d y d f ( y ∗ ) ∣ = ( y ∗ ) 3 5 = 5 1
Thus, we have for large n
∣ e n ∣ 2 ∣ e n + 1 ∣ = 5 1
so a = 1 , b = 5 , c = 2 and a + b + c = 8
Problem Loading...
Note Loading...
Set Loading...
Note that y n + 1 − 5 = 2 y n 1 ( y n − 5 ) 2 from which is it easy to show that y n converges to 5 . Note that e n = x n − φ = 2 1 ( y n − 5 ) . Thus e n 2 e n + 1 = ( y n − 5 ) 2 2 ( y n + 1 − 5 ) = y n 1 → 5 1 n → ∞ making the answer 1 + 5 + 2 = 8 .