In mathematics, particularly in computer algebra, Abramov's algorithm computes all rational solutions of a linear recurrence equation with polynomial coefficients. The algorithm was published by Sergei A. Abramov in 1989.[1][2]

Universal denominator

edit

The main concept in Abramov's algorithm is a universal denominator. Let   be a field of characteristic zero. The dispersion   of two polynomials   is defined as where   denotes the set of non-negative integers. Therefore the dispersion is the maximum   such that the polynomial   and the  -times shifted polynomial   have a common factor. It is   if such a   does not exist. The dispersion can be computed as the largest non-negative integer root of the resultant  .[3][4] Let   be a recurrence equation of order   with polynomial coefficients  , polynomial right-hand side   and rational sequence solution  . It is possible to write   for two relatively prime polynomials  . Let   and where   denotes the falling factorial of a function. Then   divides  . So the polynomial   can be used as a denominator for all rational solutions   and hence it is called a universal denominator.[5]

Algorithm

edit

Let again   be a recurrence equation with polynomial coefficients and   a universal denominator. After substituting   for an unknown polynomial   and setting   the recurrence equation is equivalent to As the   cancel this is a linear recurrence equation with polynomial coefficients which can be solved for an unknown polynomial solution  . There are algorithms to find polynomial solutions. The solutions for   can then be used again to compute the rational solutions  .[2]

algorithm rational_solutions is
    input: Linear recurrence equation  .
    output: The general rational solution   if there are any solutions, otherwise false.

     
     
     
    Solve   for general polynomial solution  
    if solution   exists then
        return general solution  
    else
        return false
    end if

Example

edit

The homogeneous recurrence equation of order   over   has a rational solution. It can be computed by considering the dispersion This yields the following universal denominator: and Multiplying the original recurrence equation with   and substituting   leads to This equation has the polynomial solution   for an arbitrary constant  . Using   the general rational solution is for arbitrary  .

References

edit
  1. ^ Abramov, Sergei A. (1989). "Rational solutions of linear differential and difference equations with polynomial coefficients". USSR Computational Mathematics and Mathematical Physics. 29 (6): 7–12. doi:10.1016/s0041-5553(89)80002-3. ISSN 0041-5553.
  2. ^ a b Abramov, Sergei A. (1995). "Rational solutions of linear difference and q -difference equations with polynomial coefficients". Proceedings of the 1995 international symposium on Symbolic and algebraic computation - ISSAC '95. pp. 285–289. doi:10.1145/220346.220383. ISBN 978-0897916998. S2CID 15424889.
  3. ^ Man, Yiu-Kwong; Wright, Francis J. (1994). "Fast polynomial dispersion computation and its application to indefinite summation". Proceedings of the international symposium on Symbolic and algebraic computation - ISSAC '94. pp. 175–180. doi:10.1145/190347.190413. ISBN 978-0897916387. S2CID 2192728.
  4. ^ Gerhard, Jürgen (2005). Modular Algorithms in Symbolic Summation and Symbolic Integration. Lecture Notes in Computer Science. Vol. 3218. doi:10.1007/b104035. ISBN 978-3-540-24061-7. ISSN 0302-9743.
  5. ^ Chen, William Y. C.; Paule, Peter; Saad, Husam L. (2007). "Converging to Gosper's Algorithm". arXiv:0711.3386 [math.CA].
  WikiProject Mathematics on Wikidata