Assignment # 12

Fibonnaci and Phi



Let's look at our sequence for the first 30 terms. As we can see, the ratio of subsequent terms approaches The Golden Ratio closer and closer.

1
1 1
2 2
3 1.5
5 1.66666666666667
8 1.6
13 1.625
21 1.61538461538462
34 1.61904761904762
55 1.61764705882353
89 1.61818181818182
144 1.61797752808989
233 1.61805555555556
377 1.61802575107296
610 1.61803713527851
987 1.61803278688525
1597 1.61803444782168
2584 1.61803381340013
4181 1.61803405572755
6765 1.61803396316671
10946 1.6180339985218
17711 1.61803398501736
28657 1.6180339901756
46368 1.61803398820532
75025 1.6180339889579
121393 1.61803398867044
196418 1.61803398878024
317811 1.6180339887383
514229 1.61803398875432

What is the Golden Ratio?

 Phi, or the Golden Ratio, is equal to the following constant:

From this, we can see that the successive ratios are getting closer and closer to the true value. Let's take a look at the variance from the actual value as well as the ratio of our values to the true value.

Fibonnaci f(n-1)/f(n) variance ratio
1
1 1 -0.618033988749895 0.618033988749895
2 2 0.381966011250105 1.23606797749979
3 1.5 -0.118033988749895 0.927050983124842
5 1.66666666666667 0.0486326779167718 1.03005664791649
8 1.6 -0.0180339887498948 0.988854381999832
13 1.625 0.0069660112501051 1.00430523171858
21 1.61538461538462 -0.00264937336527948 0.998362597211369
34 1.61904761904762 0.00101363029772417 1.00062645797602
55 1.61764705882353 -0.000386929926365465 0.999760864154242
89 1.61818181818182 0.000147829431923263 1.00009136361347
144 1.61797752808989 -0.000056460660007307 0.999965105393088
233 1.61805555555556 0.0000215668056606777 1.00001332901893
377 1.61802575107296 -8.23767693347577E-06 0.999994908835667
610 1.61803713527851 3.14652861965747E-06 1.00000194466163
987 1.61803278688525 -1.20186464891425E-06 0.999999257206797
1597 1.61803444782168 4.59071787028975E-07 1.00000028372197
2584 1.61803381340013 -1.75349769593325E-07 0.999999891627882
4181 1.61803405572755 6.69776591966098E-08 1.00000004139447
6765 1.61803396316671 -2.55831884565794E-08 0.99999998418872
10946 1.6180339985218 9.77190839357434E-09 1.00000000603937
17711 1.61803398501736 -3.73253694618825E-09 0.999999997693165
28657 1.6180339901756 1.4257022229458E-09 1.00000000088113
46368 1.61803398820532 -5.44569944693762E-10 0.999999999663437
75025 1.6180339889579 2.08007167046276E-10 1.00000000012856
121393 1.61803398867044 -7.94517784896698E-11 0.999999999950896
196418 1.61803398878024 3.03477243335237E-11 1.00000000001876
317811 1.6180339887383 -1.15918386001113E-11 0.999999999992836
514229 1.61803398875432 4.42756942220512E-12 1.00000000000274

We can also see that our sequence is oscillating around the correct value. Let's examine how this functions behaves:

 

 By the 8th term, we are pretty much constant ...

Looking at terms other than f(n+1)/f(n) ...

Fibonnaci f(n+1)/f(n) f(n+2)/f(n) f(n+3)/f(n) f(n+4)/f(n)
1
1 1 2 3 5
2 2 3 5 8
3 1.5 2.5 4 6.5
5 1.66666666666667 2.66666666666667 4.33333333333333 7
8 1.6 2.6 4.2 6.8
13 1.625 2.625 4.25 6.875
21 1.61538461538462 2.61538461538462 4.23076923076923 6.84615384615385
34 1.61904761904762 2.61904761904762 4.23809523809524 6.85714285714286
55 1.61764705882353 2.61764705882353 4.23529411764706 6.85294117647059
89 1.61818181818182 2.61818181818182 4.23636363636364 6.85454545454545
144 1.61797752808989 2.61797752808989 4.23595505617978 6.85393258426966
233 1.61805555555556 2.61805555555556 4.23611111111111 6.85416666666667
377 1.61802575107296 2.61802575107296 4.23605150214592 6.85407725321888
610 1.61803713527851 2.61803713527851 4.23607427055703 6.85411140583554
987 1.61803278688525 2.61803278688525 4.23606557377049 6.85409836065574
1597 1.61803444782168 2.61803444782168 4.23606889564336 6.85410334346505
2584 1.61803381340013 2.61803381340013 4.23606762680025 6.85410144020038
4181 1.61803405572755 2.61803405572755 4.23606811145511 6.85410216718266
6765 1.61803396316671 2.61803396316671 4.23606792633341 6.85410188950012
10946 1.6180339985218 2.6180339985218 4.23606799704361 6.85410199556541
17711 1.61803398501736 2.61803398501736 4.23606797003472 6.85410195505207
28657 1.6180339901756 2.6180339901756 4.23606798035119 6.85410197052679
46368 1.61803398820532 2.61803398820533 4.23606797641065 6.85410196461598
75025 1.6180339889579 2.6180339889579 4.2360679779158 6.85410196687371
121393 1.61803398867044 2.61803398867044 4.23606797734089 6.85410196601133
196418 1.61803398878024 2.61803398878024 4.23606797756049
317811 1.6180339887383 2.6180339887383
514229 1.61803398875432

We can see that taking the f(n+2)/f(n) sequence yields us ... phi + 1. The other terms give us successive powers of phi, i.e.:

 

This pattern is bound to continue ...

Looking starting points other than f(1) = 1 and f(2) = 1 ...

Fibonnaci f(n+1)/f(n)
1
3 3
4 1.33333333333333
7 1.75
11 1.57142857142857
18 1.63636363636364
29 1.61111111111111
47 1.62068965517241
76 1.61702127659574
123 1.61842105263158
199 1.61788617886179
322 1.61809045226131
521 1.61801242236025
843 1.61804222648752
1364 1.61803084223013
2207 1.61803519061584
3571 1.6180335296783
5778 1.61803416409969
9349 1.61803392177224
15127 1.61803401433308
24476 1.61803397897799
39603 1.61803399248243
64079 1.61803398732419
103682 1.61803398929446
167761 1.61803398854189
271443 1.61803398882935
439204 1.61803398871955
710647 1.61803398876149
1149851 1.61803398874547
Fibonnaci f(n+1)/f(n)
2
5 2.5
7 1.4
12 1.71428571428571
19 1.58333333333333
31 1.63157894736842
50 1.61290322580645
81 1.62
131 1.61728395061728
212 1.61832061068702
343 1.61792452830189
555 1.61807580174927
898 1.61801801801802
1453 1.61804008908686
2351 1.6180316586373
3804 1.61803487877499
6155 1.61803364879075
9959 1.61803411860276
16114 1.61803393915052
26073 1.61803400769517
42187 1.61803398151344
68260 1.61803399151397
110447 1.61803398769411
178707 1.61803398915317
289154 1.61803398859586
467861 1.61803398880873
757015 1.61803398872742
1224876 1.61803398875848
1981891 1.61803398874662
Fibonnaci f(n+1)/f(n)
7
59 8.42857142857143
66 1.11864406779661
125 1.89393939393939
191 1.528
316 1.6544502617801
507 1.60443037974684
823 1.6232741617357
1330 1.61603888213852
2153 1.6187969924812
3483 1.6177426846261
5636 1.61814527706001
9119 1.6179914833215
14755 1.61805022480535
23874 1.61802778719078
38629 1.61803635754377
62503 1.61803308395247
101132 1.61803433435195
163635 1.61803385674168
264767 1.61803403917255
428402 1.61803396949016
693169 1.61803399610646
1121571 1.61803398593994
1814740 1.6180339898232
2936311 1.61803398833993
4751051 1.61803398890649
7687362 1.61803398869008
12438413 1.61803398877274
20125775 1.61803398874117

It looks like it doesn't matter which values we choose as our starting point - the sequence is bound to converge to phi sooner or later. Let's look at why that's the case ...

a
b
a+b
a + 2b
2a + 3b
3a + 5b
5a + 8b
8a + 13b
13a + 21b
21a + 34b
34a + 55b
etc É

The coefficients of both the 'a' and 'b' terms are consecutive numbers from the Fibonnaci sequence. Dividing f(n+1)/f(n) will start approaching phi. Why is that the case?

Here goes our proof ...

 

 
 
 
 
 
 

The positive value will give us Phi.


Return to Pawel's 6680 Class Page