Zum Inhalt springen

Smarandache-Funktion

aus Wikipedia, der freien Enzyklopädie

In der Mathematik ist die Smarandache-Funktion eine Folge bzw. eine zahlentheoretische Funktion, die mit der Fakultät verwandt ist. Historisch gesehen wurde sie zuerst von Édouard Lucas<ref>E. Lucas: Question Nr. 288. In: Mathesis, 3, 1883, S. 232</ref> (1883), Joseph Neuberg<ref>J. Neuberg: Solutions de questions proposées, Question Nr. 288. In: Mathesis, 7, 1887, S. 68–69</ref> (1887) und Aubrey J. Kempner<ref>Aubrey J. Kempner: Miscellanea. In: American Mathematical Monthly, 25, 1918, S. 201–210, doi:10.2307/2972639</ref> (1918) betrachtet. 1980<ref>Florentin Smarandache: A Function in Number Theory. In: An. Univ. Timişoara, Ser. St. Mat., 18, 1980, S. 79–88. Vorlage:ArXiv</ref> wurde sie von Florentin Smarandache „wiederentdeckt“.

Definition und Werte

Die Smarandache-Funktion <math>\mu(n)</math> ist definiert als die kleinste natürliche Zahl, für die <math>n</math> die Fakultät von <math>\mu(n)</math> teilt.

Formal ist <math>\mu(n)</math> also die kleinste natürliche Zahl, für die gilt

<math>n\; |\; \mu(n)!</math>

Beispiele

Ist zum Beispiel der Wert <math>\mu(8)</math> gesucht, ist die kleinste der Zahlen 1!, 2!, 3!, … zu suchen, die durch 8 teilbar ist. Da <math>\, 1!=1</math> und <math>2!=1\cdot2=2</math> und <math>3!=1\cdot2\cdot3=6</math> nicht durch acht teilbar sind, <math>4!=1\cdot2\cdot3\cdot4=24=3\cdot8</math> aber doch, ist <math>\, \mu(8)=4</math>.

Allerdings ist etwa <math>\mu(7)=7</math>, da die Zahl 7 keine der Zahlen 1!, 2!, …, 6! teilt, während sie 7! trivialerweise teilt.

Die ersten Werte sind:<ref>Folge A002034 in OEIS</ref>

n 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29
<math>\mu(n)</math> 1 (*) 2 3 4 5 3 7 4 6 5 11 4 13 7 5 6 17 6 19 5 7 11 23 4 10 13 9 7 29

(*) Der Wert <math>\mu(1)</math> wird von manchen Autoren auch als 0 definiert.

Eigenschaften

Trivialerweise gilt

<math>\, \mu(n)\le n,</math>

da ja <math>n</math> auf jeden Fall <math>n!=n\cdot(n-1)!</math> teilt.

Ein grundlegendes Resultat ist, dass Gleichheit in der obigen Ungleichung genau für prime <math>n</math> oder <math>n=4</math> eintritt:

<math>\mu(n)=n\qquad\Leftrightarrow\qquad n \text{ prim}\quad \text{oder}\quad n=4</math>

Beweis:

<math>\Rightarrow</math>: Sei <math>\mu(n)=n</math> und <math>n</math> nicht prim. Dann ist <math>n=4</math> zu zeigen. Da <math>n</math> nicht prim ist, gibt es natürliche Zahlen <math>2\le s\ \le t < n</math> mit <math>n=st</math>. Wäre sogar <math>s<t</math>, so wäre <math>n=st|t!</math> und man erhielte den Widerspruch <math>\mu(n)\le t < n</math>. Also ist <math>s=t</math> und daher <math>n=t^2</math>. Wäre <math>2<t</math>, so folgte <math>t<2t<t^2=n</math>, also <math>t<2t\le n-1</math> und damit <math>n=t^2|t\cdot 2t|(2t)!|(n-1)!</math>, und man hätte erneut den Widerspruch <math>\mu(n)<n</math>. Daher muss <math>t=2</math> sein und es folgt <math>n=4</math>.

<math>\Leftarrow</math>: Ist <math>n</math> prim, so teilt <math>n</math> keine Zahl <math>m!</math> für <math>m<n</math>, da <math>n</math> per def. nicht in <math>m!</math> vorkommt. Daher gilt <math>\mu(n)=n</math>. <math>\mu(4)=4</math> ist klar.

Übrigens ergibt sich dadurch für <math>\pi(x)</math>, die Anzahl der Primzahlen kleinergleich <math>x</math> und der Ganzzahlfunktion:

<math>\pi(x)=-1+\sum_{k=2}^x \left\lfloor\frac{\mu(k)}k\right\rfloor</math>.

Nach Paul Erdős stimmt <math>\mu(n)</math> mit dem größten Primfaktor von <math>n</math> überein für asymptotisch fast alle <math>n</math>, d. h. die Anzahl der Zahlen kleiner gleich <math>n</math>, für die dies nicht gilt, ist o(n).

Allgemein gilt ferner

<math>\, \mu(n!)=n</math>

und

<math>\mu(n)\ge\mathrm{gpf}(n)</math>

wobei <math>\rm gpf</math> für den größten Primfaktor von <math>n</math> stehe.

Ganz allgemein gilt

<math>\mu\left(p_1^{\alpha_1}\cdot p_2^{\alpha_2}\cdot \ldots \cdot p_n^{\alpha_n}\right) = \max\left[\mu\left(p_1^{\alpha_1}\right), \mu\left(p_2^{\alpha_2}\right), \ldots,\mu\left(p_n^{\alpha_n}\right)\right]</math>

Für (gerade) vollkommene Zahlen <math>n</math> gilt außerdem (<math>k\in\N,p\text{ prim}</math>)<ref>Sebastián Martín Ruiz: Smarandache’s function applied to perfect numbers. In: Smarandache Notions Journal, Vol. 10, Frühjahr 1999, S. 114. Vorlage:ArXiv</ref>

<math>\mu(n)=\mu(2^{k-1}\cdot(2^k-1))=2^k-1=p</math>

Abwandlungen

Pseudosmarandache-Funktion

Die Pseudosmarandache-Funktion <math>Z(n)</math> ist die kleinste ganze Zahl, für die

<math>n\;\;\text{teilt}\;\;1+2+3+\cdots+Z(n),</math>

also das kleinste natürliche <math>n</math>, für das gilt

<math>n\;\;\left|\;\;\frac{Z(n)(Z(n)+1)}2\right.</math>

(siehe auch Dreieckszahl, Gaußsche Summenformel)

Die ersten Werte sind

1, 3, 2, 7, 4, 3, 6, 15, 8, 4, 10, 8, 12, 7, 5, 31, 16, 8, 18, 15, … (Folge A011772 in OEIS)

Einige Eigenschaften:<ref>R.G.E. Pinch: Vorlage:ArXiv in arXiv, 6. April 2005</ref>

  • <math>\sqrt n<Z(n)\le 2n-1</math>
  • <math>Z(n)\le n-1\qquad\text{für ungerade } n</math>
  • <math>Z(2^k)=2^{k+1}-1\, </math>
  • <math>\frac{Z(n+1)}{Z(n)}\text{ und } \frac{Z(n-1)}{Z(n)}\text{ und } \frac{Z(2n)}{Z(n)}</math> sind nach oben hin unbegrenzt
  • <math>\frac n{Z(n)}=k\;(k\in\Z,k\ge2)</math> hat unendlich viele Lösungen für <math>n</math>
  • <math>\sum_{n=1}^\infty \frac1{Z(n)^\alpha}</math> konvergiert für alle <math>\alpha>1</math>

Smarandache-Doppelfakultät-Funktion

Ersetzt man in der Definition die Fakultät durch die Doppelfakultät

<math>n!! = \begin{cases} n \cdot (n-2) \cdot (n-4)\cdot\ldots\cdot 2 & \text{für } n \text{ gerade,} \\

n \cdot (n-2) \cdot (n-4) \cdot\ldots\cdot 1 & \text{für } n \text{ ungerade,}\end{cases}</math> so ist <math>\mathrm{Sdf}(n)</math>

die kleinste natürliche Zahl, die durch <math>\mathrm{Sdf}(n)!!</math> teilbar ist.

Die ersten Werte für <math>\mathrm{Sdf}(n)</math> sind

1, 2, 3, 4, 5, 6, 7, 4, 9, 10, 11, 6, 13, 14, 5, 6, … (Folge A007922 in OEIS)

Smarandache-Funktion mit Primorial

Das Primorial (auch Primfakultät, <math>p_\#</math>) ist das Produkt der Primzahlen kleinergleich der gegebenen Zahl. Die Smarandache Near-to-Primorial Function<ref>{{#if: | {{{author}}} | Eric W. Weisstein }}: Smarandache Near-to-Primorial Function. In: MathWorld (englisch). {{#if: | {{#ifeq: {{#property:P2812}} | {{{id}}} | | {{#if: {{#property:P2812}} | {{#ifeq: 0 | 0 | }} | {{#ifeq: 0 | 0 | }} }} }} }}</ref> von <math>n</math> ist dann die kleinste Primzahl, für die <math>p_\#+1</math>, <math>p_\#-1</math> oder <math>p_\#</math> durch <math>n</math> teilbar ist.

Smarandache-Kurepa-Funktion und Smarandache-Wagstaff-Funktion

Für die Smarandache-Kurepa-Funktion <math>\mathrm{SK}(n)</math> wandle man die Fakultät nicht zur Doppelfakultät, sondern zu folgender Funktion ab:

<math>f(n)=\sum_{k=0}^{n-1} k!= 0!+1!+2!+\ldots+(n-1)!</math>

Für prime <math>p</math> ist <math>\mathrm{SK}(p)</math> analog die kleinste natürliche Zahl, sodass <math>f(\mathrm{SK}(p))</math> durch <math>p</math> teilbar ist.<ref>{{#if: | {{{author}}} | Eric W. Weisstein }}: Smarandache-Kurepa Function. In: MathWorld (englisch). {{#if: | {{#ifeq: {{#property:P2812}} | {{{id}}} | | {{#if: {{#property:P2812}} | {{#ifeq: 0 | 0 | }} | {{#ifeq: 0 | 0 | }} }} }} }}</ref>

Die ersten Werte sind 2, 4, 6, 6, 5, 7, 7, 12, 22, 16, 55 und bilden Folge A049041 in OEIS.

Die Smarandache-Wagstaff-Funktion verwendet stattdessen<ref>{{#if: | {{{author}}} | Eric W. Weisstein }}: Smarandache-Wagstaff Function. In: MathWorld (englisch). {{#if: | {{#ifeq: {{#property:P2812}} | {{{id}}} | | {{#if: {{#property:P2812}} | {{#ifeq: 0 | 0 | }} | {{#ifeq: 0 | 0 | }} }} }} }}</ref>

<math>f(n)=\sum_{k=1}^n k!= 1!+2!+\ldots+n!</math>

Smarandache-Ceil-Funktion

Die Smarandache-Wagstaff-Funktion k-ter Ordnung <math>S_k(n)</math> schließlich ist als die kleinste natürliche Zahl definiert, für die <math>\, [S_k(n)]^k</math> durch <math>n</math> teilbar ist.<ref>{{#if: | {{{author}}} | Eric W. Weisstein }}: Smarandache Ceil Function. In: MathWorld (englisch). {{#if: | {{#ifeq: {{#property:P2812}} | {{{id}}} | | {{#if: {{#property:P2812}} | {{#ifeq: 0 | 0 | }} | {{#ifeq: 0 | 0 | }} }} }} }}</ref>

Die ersten Werte:

<math>k</math> <math>S_k(n)</math>
1 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, … <math>n</math>
2 1, 2, 3, 2, 5, 6, 7, 4, 3, 10, 11, 6, 13, 14, 15, … (Folge A019554 in OEIS)
3 1, 2, 3, 2, 5, 6, 7, 2, 3, 10, 11, 6, 13, 14, 15, … (Folge A019555 in OEIS)
4 1, 2, 3, 2, 5, 6, 7, 2, 3, 10, 11, 6, 13, 14, 15, … (Folge A053166 in OEIS)

Weiteres

  • Tutescu<ref>L. Tutescu: On a Conjecture Concerning the Smarandache Function. Abstracts of Papers Presented to the American Mathematical Society 17, S. 583, 1996</ref> vermutete, dass für zwei aufeinanderfolgende Zahlen deren Werte der Smarandache-Funktion stets verschieden sind:
<math>\mu(n)\not=\mu(n+1)\qquad\quad\text{für alle } n</math>
Die Vermutung wurde bis <math>10^9</math> bestätigend nachgerechnet.
Die Reihe der Kehrwerte der Fakultäten der Smarandache-Funktion konvergiert (erste Smarandache-Konstante):
<math>\sum_{n=2}^\infty \frac1{\mu(n)!}=1{,}09317\dots</math> (Folge A048799 in OEIS)

Literatur

  • Kenichiro Kashihara: Comments and topics on Smarandache notions and problems. (PDF; 1,6 MB) Erhus University Press 1996, ISBN 1-87958-555-3.
  • Norbert Hungerbühler, Ernst Specker: A Generalisation of the Smarandache Function to Several Variables. (PDF; 220 kB) In: Electronic Journal of Combinatorical Number Theory, 6, 2006, #A23.
  • C. Dumitrescu, N. Virlan, St. Zamfir, E. Radescu, N. Radescu, F.Smarandache: Smarandache Type Function Obtained by Duality. In: Studii si Cercetari Stiintifice, Seria: Matematica, University of Bacau, No. 9, 1999, S. 49–72, Vorlage:ArXiv.
  • Sebastian Martin Ruiz, M. L. Perez: Properties and Problems related to Smarandache Type Functions. In: Mathematics Magazine for grades 1-12, 2/2004, S. 46–53, Vorlage:ArXiv.

Weblinks

Einzelnachweise

<references />