Mathematical Programming

, Volume 62, Issue 1–3, pp 15–39 | Cite as

Solving symmetric indefinite systems in an interior-point method for linear programming

  • Robert Fourer
  • Sanjay Mehrotra


We describe an implementation of a primal—dual path following method for linear programming that solves symmetric indefinite “augmented” systems directly by Bunch—Parlett factorization, rather than reducing these systems to the positive definite “normal equations” that are solved by Cholesky factorization in many existing implementations. The augmented system approach is seen to avoid difficulties of numerical instability and inefficiency associated with free variables and with dense columns in the normal equations approach. Solving the indefinite systems does incur an extra overhead, whose median is about 40% in our tests; but the augmented system approach proves to be faster for a minority of cases in which the normal equations have relatively dense Cholesky factors. A detailed analysis shows that the augmented system factorization is reliable over a fairly large range of the parameter settings that control the tradeoff between sparsity and numerical stability.

Key words

Linear programming interior-point methods symmetric indefinite systems 


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. [1]
    I. Adler, N. Karmarkar, M.G.C. Resende and G. Veiga, “An implementation of Karmarkar's algorithm for linear programming,”Mathematical Programming 44 (1989) 297–336.Google Scholar
  2. [2]
    I. Adler, N. Karmarkar, M.G.C. Resende and G. Veiga, “Data structures and programming techniques for the implementation Karmarkar's algorithm,”ORSA Journal on Computing 1 (1989) 84–106.Google Scholar
  3. [3]
    M. Arioli, I.S. Duff and P.P.M. de Rijk, “On the augmented system approach to sparse least-squares problems,”Numerische Mathematik 55 (1989) 667–684.Google Scholar
  4. [4]
    R.E. Bixby, “Implementing the simplex method: The initial basis,”ORSA Journal on Computing 4 (1992) 267–284.Google Scholar
  5. [5]
    P.T. Boggs, P.D. Domich, J.R. Donaldson and C. Witzgall, “Optimal 3-dimensional methods for linear programming,” MISTIR 89-4225, Center for Computing and Applied Mathematics, National Institute of Standards and Technology (Gaithersburg, MD, 1989).Google Scholar
  6. [6]
    Å. Björck, “Iterative refinement of linear least-squares solutions,”BIT 7 (1967) 257–278.Google Scholar
  7. [7]
    J.R. Bunch and B.N. Parlett, “Direct methods for solving symmetric indefinite systems of linear equations,”SIAM Journal on Numerical Analysis 8 (1971) 639–655.Google Scholar
  8. [8]
    I.C. Choi, C.L. Monma and D.F. Shanno, “Further development of a primal—dual interior point method,”ORSA Journal on Computing 2 (1990) 304–311.Google Scholar
  9. [9]
    I.S. Duff, N.I.M. Gould, J.K. Reid, J.A. Scott and K. Turner, “The factorization of sparse symmetric indefinite matrices,”IMA Journal of Numerical Analysis 11 (1991) 181–204.Google Scholar
  10. [10]
    I.S. Duff and J.K. Reid, “The multifrontal solution of indefinite sparse symmetric linear equations,”ACM Transactions on Mathematical Software 9 (1983) 302–325.Google Scholar
  11. [11]
    R. Fourer and S. Mehrotra, “Solving Symmetric Indefinite Systems in an Interior-Point Method for Linear Programming,” Technical Report 92-01, Department of Industrial Engineering and Management Sciences, Northwestern University (Evanston, IL, 1992).Google Scholar
  12. [12]
    D.M. Gay, “Electronic mail distribution of linear programming test problems,”Committee on Algorithms Newsletter 13 (1985) 10–12.Google Scholar
  13. [13]
    P.E. Gill, W. Murray, D.B. Ponceleón and M.A. Saunders, “Solving reduced KKT systems in barrier methods for linear and quadratic programming,” Technical Report SOL 91-7, Systems Optimization Laboratory, Stanford University (Stanford, CA, 1991).Google Scholar
  14. [14]
    P.E. Gill, W. Murray, M.A. Saunders, J.A. Tomlin and M.H. Wright, “On projected Newton barrier methods for linear programming,”Mathematical Programming 36 (1986) 183–209.Google Scholar
  15. [15]
    N. Karmarkar, “A new polynomial time algorithm for linear programming,”Combinatorica 4 (1984) 373–393.Google Scholar
  16. [16]
    M. Kojima, S. Mizuno and A. Yoshise, “A primal—dual interior point algorithm for linear programming,” in: N. Megiddo, ed.,Progress in Mathematical Programming: Interior-Point and Related Methods (Springer, New York, 1989) pp. 131–158.Google Scholar
  17. [17]
    I.J. Lustig, R.E. Marsten and D.F. Shanno, “On implementing Mehrotra's predictor-corrector interior-point method for linear programming,”SIAM Journal on Optimization 2 (1992) 435–449.Google Scholar
  18. [18]
    I.J. Lustig, J.M. Mulvey and T.J. Carpenter, “Formulating two-stage stochastic programs for interior point methods,”Operations Research 39 (1991) 757–770.Google Scholar
  19. [19]
    H.M. Markowitz, “The elimination form of the inverse and its application to linear programming,”Management Science 3 (1957) 255–269.Google Scholar
  20. [20]
    K.A. McShane, C.L. Monma and D.F Shanno, “An implementation of a primal—dual interior point method for linear programming,”ORSA Journal on Computing 1 (1989) 70–83.Google Scholar
  21. [21]
    S. Mehrotra, “On the implementation of a primal—dual interior point method,”SIAM Journal on Optimization 2 (1992) 575–601.Google Scholar
  22. [22]
    S. Mehrotra, “Higher order methods and their performance,” Technical Report 90-16R1, Department of Industrial Engineering and Management Sciences, Northwestern University (Evanston, IL, 1991).Google Scholar
  23. [23]
    S. Mehrotra, “Handling free variables in interior methods,” Technical Report 91-06, Department of Industrial Engineering and Management Sciences, Northwestern University (Evanston, IL, 1991).Google Scholar
  24. [24]
    S. Mehrotra and Y. Ye, “On finding the optimal facet of linear programs,” Technical Report 91-10. Department of Industrial Engineering and Management Sciences, Northwestern University (Evanston, IL, 1991).Google Scholar
  25. [25]
    R.D.C. Monteiro and I. Adler, “Interior path following primal—dual algorithms. Part I: Linear programming,”Mathematical Programming 44 (1989) 27–41.Google Scholar
  26. [26]
    R.D. Skeel, “Scaling for numerical stability in Gaussian elimination,”Journal of the Association for Computing Machinery 26 (1979) 494–526.Google Scholar
  27. [27]
    K. Turner, “Computing projections for the Karmarkar algorithm,”Linear Algebra and its Applications 152 (1991) 141–154.Google Scholar
  28. [28]
    R.J. Vanderbei, “Splitting dense columns in sparse linear systems,” Manuscript, AT&T Bell Laboratories (Murray Hill, NJ, 1990).Google Scholar
  29. [29]
    R.J. Vanderbei, “ALPO: another linear program solver,” Manuscript, AT&T Bell Laboratories (Murray Hill, NJ, 1990).Google Scholar
  30. [30]
    R.J. Vanderbei and T.J. Carpenter, “Symmetric indefinite systems for interior-point methods,” Technical Report SOR-91-7, Department of Civil Engineering and Operations Research, Princeton University (Princeton, NJ, 1991).Google Scholar

Copyright information

© The Mathematical Programming Society, Inc. 1993

Authors and Affiliations

  • Robert Fourer
    • 1
  • Sanjay Mehrotra
    • 1
  1. 1.Department of Industrial Engineering and Management SciencesNorthwestern UniversityEvanstonUSA

Personalised recommendations