Abstract
We show that the problem of accumulating Jacobian matrices by using a minimal number of floating-point operations is NP-complete by reduction from Ensemble Computation. The proof makes use of the fact that, deviating from the state-of-the-art assumption, algebraic dependences can exist between the local partial derivatives. It follows immediately that the same problem for directional derivatives, adjoints, and higher derivatives is NP-complete, too.
Similar content being viewed by others
References
Personal communication with Steihaug, T. Bergen University, Norway, and Griewank, A. at Humboldt University Berlin (2005)
Baur W., Strassen V. (1983) The complexity of partial derivatives. Theoret. Comput. Sci. 22, 317–330
Berz, M., Bischof, C., Corliss, G., Griewank, A. (eds.) Computational differentiation: techniques, applications, and tools. In: Proceedings Series. SIAM Philadelphia (1996)
Bischof, C., Haghighat, M.: Hierarchical approaches to automatic differentiation. In: [3], pp. 82–94
Bücker, M., Corliss, G., Hovland, P., Naumann, U., Norris, B. (eds.) Automatic Differentiation: Applications, Theory, and Tools. Lecture Notes in Computational Science and Engineering, vol. 50. Springer, Berlin Heidelberg New York (2005)
Corliss G., Faure C., Griewank A., Hascoët L., Naumann U. (eds) (2002) Automatic Differentiation of Algorithms – From Simulation to Optimization. Springer, Berlin Heidelberg New York
Corliss, G., Griewank, A. (eds.) Automatic Differentiation: Theory, Implementation, and Application. Proceedings Series. SIAM Philadelphia (1991)
Garey M., Johnson D. (1979) Computers and Intractability – A Guide to the Theory of NP-completeness. W. H. Freeman and Company, San Francisco
Giering, R., Kaminski, T.: Applying TAF to generate efficient derivative code of Fortran 77-95 programs. In: Proceedings of GAMM 2002, Augsburg, Germany (2002)
Griewank A. (2000) Evaluating derivatives. Principles and techniques of algorithmic differentiation Frontiers in Applied Mathematics, vol 19. SIAM, Philadelphia
Griewank A., Juedes D., Utke J. (1996) ADOL–C, a package for the automatic differentiation of algorithms written in C/C++. ACM Trans. Math. Softw. 22(2): 131–167
Griewank A., Naumann U. (2003) Accumulating Jacobians as chained sparse matrix products. Math. Prog. 3(95): 555–571
Griewank, A., Reese, S.: On the calculation of Jacobian matrices by the Markovitz rule. In: [7], pp. 126–135
Griewank, A., Vogel, O.: Analysis and exploitation of Jacobian scarcity. In: Proceedings of HPSC Hanoi. Springer, Berlin Heidelberg New York (2003)
Hascoët, L., Pascual, V.: Tapenade 2.1 user’s guide. Technical report 300, INRIA (2004)
Heath M. (1998) Scientific Computing. An Introductory Survey. McGraw-Hill, New York
Kelley, C.: Solving Nonlinear Equations with Newton’s Method. SIAM Philadelphia (2003)
Naumann, U.: Elimination techniques for cheap Jacobians. In: [6], chap. 29, pp. 247–253 (2001)
Naumann U. (2002) Cheaper Jacobians by simulated annealing. SIAM J. Opt. 13(3): 660–674
Naumann U. (2004) Optimal accumulation of Jacobian matrices by elimination methods on the dual computational graph. Math. Prog. 3(99): 399–421
Naumann U., Utke J. (2005) Optimality-preserving elimination of linearities in Jacobian accumulation. Electron. Trans. Numer. Anal. (ETNA) 21, 134–150
Naumann, U., Utke, J., Wunsch, C., Hill, C., Heimbach, P., Fagan, M., Tallent, N., Strout, M.: Adjoint code by source transformation with open ad/f. In: Proceedings of the European Conference on Computational Fluid Dynamics (ECCOMAS CFD 2006). TU Delft (2006)
Speelpenning, B.: Compiling fast partial derivatives of functions given by algorithms. Ph.D. Thesis, University of Chicago (1980)
Walther, A.: Program reversal schedules for single- and multi-processor machines. Ph.D. Thesis, Institute of Scientific Computing, Technical University Dresden (1999)
Walther, A., Griewank, A.: New results on program reversals. In: [6], chapt. 28, pp. 237–243. Springer, Berlin Heidelberg New York (2001)
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Naumann, U. Optimal Jacobian accumulation is NP-complete. Math. Program. 112, 427–441 (2008). https://doi.org/10.1007/s10107-006-0042-z
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10107-006-0042-z