Discrete & Computational Geometry

, Volume 5, Issue 1, pp 13–26 | Cite as

Geometry of the Gass-Saaty parametric cost LP algorithm

  • Victor Klee
  • Peter Kleinschmidt


In previous discussions of the Gass-Saaty algorithm, the possibility of cycling is avoided by making strong nondegeneracy assumptions or by incorporating a lexicographic decision rule. By analyzing the geometric ideas on which the algorithm is based, it is shown here that even without any “lexicography,” cycling is impossible unless the two objective functions are related in a very special way to each other or to the constraints defining the feasible regionP. In particular, the avoidance of cycling does not require any restriction on the facial structure ofP or on the algebraic relationships among the linear equalities and inequalities by means of whichP is defined.


Feasible Region Discrete Comput Geom Linear Functional Simplex Algorithm Polyhedral Cone 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.


  1. [1]
    I. Adler, R. Karp, and R. Shamir, A simplex variant solving anm × d linear program inO(min(m 2,d 2)) expected number of pivots,J. Complexity 3 (1987), 372–387.MATHMathSciNetCrossRefGoogle Scholar
  2. [2]
    K.-H. Borgwardt, Untersuchungen zur Asymptotik der mittleren Schrittzahl von Simplex-verfahren in der linearen Optimierung,Oper. Res. Verf. 28 (1978), 333–345.Google Scholar
  3. [3]
    K.-H. Borgwardt, Some distribution-independent results about the asymptotic order of the average number of pivot steps of the simplex method,Math. Oper. Res. 7 (1982), 441–462.MathSciNetCrossRefGoogle Scholar
  4. [4]
    K.H. Borgwardt, The average number of pivot steps required by the simplex-method is polynomial,Z. Oper. Res. Ser. A-B 26 (1983), A157-A177.MathSciNetGoogle Scholar
  5. [5]
    A. Gana, Studies in the complementarity problem, Ph.D. Dissertation, Department of Industrial and Operations Engineering, University of Michigan, Ann Arbor (1982).Google Scholar
  6. [6]
    S. Gass,Linear Programming, McGraw-Hill, New York, 1975.Google Scholar
  7. [7]
    S. Gass and T. Saaty, The computational algorithm for the parametric objective function,Naval Res. Logist. Quart. 2 (1955), 39–45.MathSciNetCrossRefGoogle Scholar
  8. [8]
    B. Grünbaum,Convex Polytopes, Wiley-Interscience, London, 1967.MATHGoogle Scholar
  9. [9]
    M. Haimovich, The simplex algorithm is very good! On the expected number of pivot steps and relatived properties of random linear programs, Unpublished manuscript, Columbia University, New York (1983).Google Scholar
  10. [10]
    K. Murty, Computational complexity of parametric linear programming,Math. Programming 19 (1980), 213–219.MathSciNetCrossRefGoogle Scholar
  11. [11]
    K. Murty,Linear Programming, Wiley, New York, 1983.MATHGoogle Scholar
  12. [12]
    A. Schrijver,Theory of Linear and Integer Programming, Wiley-Interscience, New York, 1986.MATHGoogle Scholar

Copyright information

© Springer-Verlag New York Inc. 1990

Authors and Affiliations

  • Victor Klee
    • 1
  • Peter Kleinschmidt
    • 2
  1. 1.Department of MathematicsUniversity of WashingtonSeattleUSA
  2. 2.Institut für Mathematik und InformatikUniversität PassauPassauGermany

Personalised recommendations