Composite Number Sequence Challenge 1/97

by

Huen Y.K.

CAHRC, P.O.Box 1003, Singapore 911101
http://web.singnet.com.sg/~huens/
email: huens@mbox3.singnet.com.sg

(A short communication - 1st released: 28/6/97. Revised 28/6,29/6,2/7)


Abstract

The mist is beginning to lift on the properties of compositeness and primeness of natural numbers. It is very easy to develop closed formulations for divisibility sequences but not primeness sequences. In sequence algebra, one can still develop global, explicit formulations for difficult sequences like Mersenne primes and Fermat's primes but the use of mixed mode formulations is unavoidable and this has to do with the presence of unnormalised sequences. There is no direct, unique tests for indivisibility but there exists quite a few probability primality tests of great ingenuity and utilities. Anyone wishing to formulate a difficult prime sequence should make sure that the intervals between successive primes grow big very fast and nonlinearly. Not much interest has be paid to developing difficult composite number sequences and yet the same "trade secrets" can be applied. This paper describes the techniques in devising difficult sequences and poses a fairly easy one as a first challenge. Sorry, there is no monetary prize for it but the author will issue nice multi-coloured laser printed certificates for the winner and two runners up. The closing date is at 12 noon on 31.8.97. Please submit your results by email to huens@singnet.com.sg or huens@mbox3.singnet.sg.




1. Introduction

Readers who are only interested in the challenge problem are requested to proceed directly to section (4) of this paper.

Given a finite sequence of integers, derive a deterministic, closed global formulation which can predict all subsequent members of the sequence. Some sequence problems are posted in A.T. & T. Integer Research Website but those do not require finding any closed forms [3].

With the development of sequence algebra since 1994, finding determinstic formulations for Mersenne primes and Fermat primes has become quite straightforward provided either mixed mode arithmetic or the use of Normc( ) operator is allowed. This paper explains how exotic composite sequences can be developed which can rival the difficulties encountered in the prediction of prime sequences such as Mersenne prime and Fermat prime sequences. Note that the author refers to these as sequences rather than as Mersenne and Fermat numbers [16]. For demonstration of the algebraic techniques involved, the author will describe how these two formulations are developed. Then attention will be shifted to the development of exotic composite sequences. The latter will have more recreational values since the common starting point is the general equation of divisibles which by now should be familiar to visitors to this URL-site.


2. Mersenne Primes And Fermat Primes

(i) Explicit Formulation For Mersenne Prime Sequence: Merzprime(z)

Even now number theoretic hunters tend to concentrate on individual Mersenne primes and Fermat primes instead of their holistic sequences. There is a Mersenne Prime Webpage hosted by Chris K. Caldwell which reports the latest findings on Mersenne primes and Perfect numbers amongst other interesting historical facts [2]. Until 1996, no determinstic algebraic formulations have been developed for Mersenne primes or Fermat primes[17]. There exists a defintion for Mersenne prime which offers meagre help to those in the Mersenne hunt. It is as much help to a blue whale hunter by telling him that this is the largest creature on planet earth and that it is somewhere in the vast ocean.

.............Definition: When 2^n -1 is prime, it is said to be a Mersenne prime.

If a coincidence between a prime in the Mersenne prime sequence and one in the natural prime sequence is defined as a hit, then the hit rate for Mersenne primes is very low indeed and for Fermat primes even lower. This can be easily demonstrated by a simple Maple program line with embedded primality tests. Such primality tests are probablistic but so far no discovery of false primes have been reported. Primality tests are convenient but sequence algebraists have a superior weapon -- a determinstic global formulation for the prime sequence called Prime(z) [13]. Equation (1) shows the Laurent series of the first 100 terms of a Mersenne number sequence generated using primality tests on the integer i. All Mersenne primes are displayed in bold fonts.

Even amongst the first 100 integers, the hit rate is already quite low but we know that intervals between successive Mersenne primes grow very fast with large integers. This explains why it is so difficult to break world records in the Mersenne hunt [2]. This is because determinstic primality tests on large integers will get increasingly arduous and the rarity of large Mersenne primes probably turns this into a game of patience. It looks as if the Mersenne hunt will in future be the reserve of big game hunters, i.e., those in possession of mainframes. The author recommends sequence algebraic games since these do not require immense number crunching power.


Mersenneprime(z):=sum(x^i/z^(2^i-1)*(isprime(i)-false)/(true-false),i=2..100);

......................................2.......3.......5.......7.........11.........13..........17...........19..............23
....................................x.......x.......x.......x......... x...........x...........x..............x...............x
Mersennenumb(z) := ---- + ---- + --- + ---- + ------ + ------ + -------- + --------- + ----------
......................................3........7......31....127....2047.....8191......131071....524287.....8388607
....................................z........z.......z.......z.........z............z............z..............z...............z

...........29.............31................37.......................41............................43.............................47
.........x...............x...................x.........................x..............................x...............................x
+ -----------+ -------------+ ---------------+ ------------------+ -----------------+ ----------
...536870911..2147483647..137438953471..2199023255551..8796093022207..140737488355327 z....................z.....................z........................z............................z..........................z

...............53..........................59.....................................61.................................67
..............x...........................x.......................................x...................................x
+ -------------------+ -----------------------+ --------------------------+ -----------------
9007199254740991 576460752303423487 2305843009213693951 147573952589676412927
z...............................z...................................z.......................................z

.....................71..........................................73........................................79
...................x............................................x..........................................x
+ --------------------------------- + --------------------------------- + -------------------------
...........2361183241434822606847....9444732965739290427391....604462909807314587353087
..........z.............................................z..............................................z

......................83.............................................89
....................x...............................................x
+ ------------------------------- + --------------------------------- ..9671406556917033397649407....618970019642690137449562111
z...................................................z

.....................97
...................x
+ -------------------------------
....158456325028528675187087900671 ...................................................................(1).
..z

We proceed to develop a determinstic formulation for Mersenne primes as follows:

Nat(z):=series(z/(z-1),z=infinity,10);

..........................................1..........1.......1........1.......1.......1.........1.........1............1
..........Nat(z) := 1 + 1/z + ---- + ---- + ---- + ---- + ---- + ---- + ---- + ---- + O(---) .......(2)
............................................2..........3.......4........5........6.......7........8..........9..........10
..........................................z...........z.......z........z.........z.......z........z..........z.............z

Comp(z):=series(sum(1/(z^i*(z^i-1)),i=2..10),z=infinity,10);

................................................1.......2.........2.........1............1
............................Comp(z) := ---- + ---- + ---- + ---- + O(---) ...................................(3).
..................................................4.......6.........8.........9...........10
.................................................z.......z.........z.........z...........z

One already notices that Nat(z) is normalised but Comp(z) is not. To proceed we need to normalise Comp(z) as follows:

.................................................1..........1........1........1...........1
.................Normc(Comp(z)) := ---- + ---- + ---- + ---- + O(---) ...................................(4).
...................................................4..........6........8........9..........10
.................................................z...........z........z.........z..........z

Now Prime(z) can be derived by the difference between Nat(z) and Normc(Comp(z)) as follows:

Prime(z):= Nat(z) - Normc(Comp(z)) - 1 - 1/z

.............................1......1........1........1........1.......1.......1.......1.......1.......1......1
........Prime(z) := ---- + ---- + ---- + ---- + --- + --- + --- + --- + --- + --- + --- ................(5).
...............................2......3........5........7.......11......13......17.....19.....23.....29.....31
.............................z.......z........z.........z........z.........z........z.......z.......z........z.......z

Merznumb(z):=sum(1/z^(2^i-1),i=2..5);

..................................................1........1......1........1
......................Merznumb(z) := ---- + ---- + --- + ----... ................................(6).
....................................................3........7......15......31
..................................................z.........z.......z.......z

By applying logic AND between equations (5) and (6) Mersenne primes can be filtered out as shown in equation (7). Even though a short example is shown here, this is a global formulation for Mersenne primes.

Merzprime(z):= [AND Prime(z) Merznumb(z)] :=

...................................................1........1.........1
.......................Merzprime(z) := ---- + ---- + ------- +... ................................(7).
.....................................................3........7........31
...................................................z.........z........z

(ii) Explicit Formulation For Fermat prime sequence: Fermatprime(z)

Fermat primes are harder to compute since these grow big even faster and the intervals between successive Fermat primes are even more nonlinear. Nevertheless the techniques developed for Mersenne prime sequence can be applied as follows:

Prime(z):= Nat(z) - Normc(Comp(z)) - 1 - 1/z

...............................1........1.......1......1.......1.......1.........1
.........Prime(z) := ---- + ---- + ---- + ---- + --- + --- + --- + .....
.................................2.........3.......5......7......11......13.......17
...............................z..........z........z.......z.......z........z.........z

...............................1........................1....................................1
............................. ---- +...... + -------------- + .........+ ---------------
.................................257...................65537..........................4294967207 ............(8).
...............................z.......................z...................................z

Fermatnumb(z):=sum(1/z^(2^(2^i)+1),i=2..5);

.......................................................1........1...........1.................1
.........................Fermatnumb(z) := ----- + ------ + -------- + -------------
........................................................17......257........65537.........4294967297 ........(9).
.......................................................z........z............z.................z

.. By applying logic AND between equations (8) and (9) the Mersenne primes will be filtered out as shown in equation (7). Even though a short example is shown here, equation (9) is a global deterministic formulation for Fermat primes since it does not make use of primality tests.

Fermatprime(z):= [AND Prime(z) Fermatnumb(z)] :=

.......................................................1.............1.............1
........................Fermatprime(z) := ------ + ------- + --------- ..................(10).
........................................................17..........257...........65537
.......................................................z............z..............z

It can be seen that Fermatprimes(z) gets exceedingly rare as integers move into the very large end.


3. Exotic Composite Sequences

All exotic composite sequences has its starting point from the generalised equation of divisibles given by equation (11):

..................................................................1
...................................Gencomp(z) := ----------------- .........................................(11).
.............................................................f(i).....f(i)
............................................................z.....(z.... - 1)

In the case of the natural composite sequence Comp(z), f(i) = i is linear. All one needs to do is to find functions of f(i) which are nonlinear and the magnitudes of which grow very fast. The condition which is imposed in the challenge problem posed in this paper is that f(i) must compute to positive integers. A demonstration examples is given here in equation (12). The present problem is still quite mild since intervals are purposedly kept small.

Gencomp(z):= series(sum(1/(z^(i^2+i+1)*(z^(i^2+i+1)-1)),i=2..90),z=infinity,90);

..........................1.......1.......1......1......1.......1.......2......1......1......1......1.......2.......1......1......1
Gencomp(z) := --- + --- + --- + --- + --- + --- + --- + --- + --- + --- + --- + --- + --- + -- + --
..........................14.....21....26.....28....35.....39......42.....49.....52....56....62.....63.....65....70....77 .........................z.......z.......z........z.......z........z........z......z.......z......z......z.......z.......z.......z.......z

.............................1.......2.......1...........1
.........................+ --- + --- + --- + O(---) .......................................................... (12).
..............................78......84......86........91
.............................z.......z.........z..........z

Thus the demo-problem posed will be as a normalised sequence given by the integer set shown in equation (13). The challenge will be to find the closed formulation f(i) for this composite sequence. Of course the correct answer is already given in equation (12) as i^2+i+1.

{14 21 26 28 35 39 42 49 52 56 62 63 65 70 77 78 84 86 91} ......(13).


4. The Challenge Problem Called Comp1_97(z)

The announcement is made of a challenge problem called Comp1_97(z) in this webpage. This paper can be found under section (8) of this webpage. All subsequent challenge problem for composite sequences will bear the label Compn_97(z) for the year 1997. Here is the first challenge problem:

Comp1_97(z): Find the close sequence algebraic formulation f(i) for the following composite integer set. Note that Comp1_97(z) is defined to take the same upperbound ub for both the parameters i and z.

Comp1_97(z):= series(sum(1/(z^f(i)*(z^f(i)-1)),i=2..ub),z=infinity,ub);

............{ 22 33 44 55 62 66 77 88 93 99 110 121 }

Please just email the f(i) expression to the author at address below. If your submission is received you will get an acknowledgement by return email. If you do not hear from the organiser, please resubmit another one within 24 hours.

.............huens@singnet.com.sg or huens@mbox3.singnet.com.sg

The closing date for the challenge is at 12.00 noon Singapore time on 31.8.97, i.e. August the 31st of 1997. The first correct submission received will get a nice Certificate of Achievement recording your findings which you could display in your living room or your workplace. The second and third correct answers will also get Certificates of Merit.

Final results will be announced in this webpage.

6. Conclusions

This paper is designed to introduce a new recreational game in composite number sequences in addition to the original game of the grand search for full Goldbach's Sequence which is still in progress. The objective is to introduce games which are more algebraic than numerical so that these are accessible to most PC owners.

7. References

1. Burton M.D. :Elementary Number Theory, Third Edition, WCB Publisher, 1994, pp 203 to 226.

2. Chris K. Caldwell : Mersenne Primes: History, Theroems and Lists. .

3. AT & T Integer Sequences Research : Sloane's On-Line Encyclopedia of Integer Sequences. URL site: http://www.research.alt.com/~njas/ sequences/ index. html

4. Huen Y.K.: A Simple Introduction To Sequence Algebra, URL site: http://web.singnet.com.sg/~huens/

5. Huen Y.K.: The Canonical Generating Function or CGF(z) - a Swiss-knife function. URL site: http://web.singnet.com.sg/~huens/ .

6. Huen Y.K.: Information Contents Of Number Theoretic Functions. URL site: http://web.singnet.com.sg/~huens/ .

7. Huen Y.K.: In Search Of Exotic Arithmetic Operators, URL site: http://web.singnet.com.sg /~huens/ .

8. Huen Y.K.: Visual Solutions Of Number Theoretic Functions in Multidimensional Sequence Space, URL site: http://web.singnet.com.sg /~huens/ .

9. Huen Y.K.: Final Value Theorems Applied To Number Sequences -- its strengths and weaknesses, URL site: http://web.singnet.com.sg /~huens/ .

10. Huen Y.K.: Unsolved Problems In Sequence Algebra, URL site: http://web.singnet.com.sg /~huens/ .

11. Huen Y.K.: Explicit Formulation For Modular Arithmetic In Sequence Algebra, URL site: http://web.singnet.com.sg /~huens/ .

12. Huen Y.K.: Cyclic Generating Functions In Sequence Algebra, URL site: http://web.singnet.com.sg /~huens/ .

13. Huen Y.K. : Methods Of Developing Sequence Algebraic Formulations For Comp(z) and Prime(z). URL site: http://web.singnet.com.sg /~huens/ .

14. Huen Y.K. : Information Contents Of Hypothetical DNA Sequences. URL site: http://web.singnet.com.sg/~huens/ .

15. Huen Y.K.: A Matrix Map for Prime and Non-prime Numbers, INT. J. Math. Educ. Sci. Technol., 1994, VOL. 25, NO.6, pp 913-920.

16. Huen Y.K.: Some Interesing Properties Of The Natural Number System, Int. J. Math. Educ. Sci. Technol., 1996, VOL.27, NO. 5, 685-691.

17. Huen Y.K.: Visual algebra and its applications, INT. J. Math. Educ. Sci. Technol.,1996, VOL.??, NO.?, ???-??? (In the press as proof paper mes 100421).

18. Huen Y.K.: Twin primes revisited: INT. J. Math. Educ. Sci. Technol., 1997, VOL.??,NO.?, ???-???. (In the press as proof paper mes 100488).

======================== END OF PAPER ======================