Automatic differentiation in Odyssée, This paper describes the design of Odyssée, a system for fortran programs manipulations and its application to automatic differentiation. The Odyssée system manipulates fortran programs as symbolic objects. It is an open system built as a toolkit, written in a high-level programming language adapted to this purpose. The use of a variational method to perform data assimilation requires the computation of the gradient of a cost function represented by a large-size fortran program. The usual drawback in the reverse automatic differentiation method is the storage requirement. The Odyssée system allows one to implement storage/recomputation strategies in order to fit the needed compromizes. We present the implementation of the strategy used in the weather forecasting arpege/ifs project to produce the adjoint code from the code representing the numerical model. Odyssée produces the same code as the hand-written adjoint code for the arpege/ifs project.

References in zbMATH (referenced in 29 articles )

Showing results 1 to 20 of 29.
Sorted by year (citations)

1 2 next

  1. Hascoet, Laurent; Pascual, Valérie: The Tapenade automatic differentiation tool, principles, model, and specification (2013)
  2. Lawless, Amos S.: Variational data assimilation for very large environmental problems (2013)
  3. Cheng, Qiang; Cao, Jianwen; Wang, Bin; Zhang, Haibin: Adjoint code generator (2009)
  4. Bischof, Christian H.; Hovland, Paul D.; Norris, Boyana: On the implementation of automatic differentiation tools (2008)
  5. Utke, Jean; Naumann, Uwe; Fagan, Mike; Tallent, Nathan; Strout, Michelle Mills; Heimbach, Patrick; Hill, Chris; Wunsch, Carl: OpenAD/F: A modular open-source tool for automatic differentiation of Fortran codes. (2008)
  6. Gebremedhin, Assefaw Hadish; Manne, Fredrik; Pothen, Alex: What color is your Jacobian? Graph coloring for computing derivatives (2005)
  7. Bischof, Christian H.; Bücker, H.Martin; Wu, Po-Ting: Time-parallel computation of pseudo-adjoints for a leapfrog scheme (2004)
  8. Griesse, R.; Walther, A.: Evaluating gradients in optimal control: continuous adjoints versus automatic differentiation (2004)
  9. Mohammadi, Bijan: Optimization of aerodynamic and acoustic performances of supersonic civil transports (2004)
  10. Daumas, Laurent; Quang Vinh Dinh; Kleinveld, Steven; Rogé, Gilbert: How to take into account deformation in a CAD-based Euler optimization process? (2003)
  11. Gunzburger, Max D.: Perspectives in flow control and optimization (2003)
  12. Martins, Joaquim R.R.A.; Sturdza, Peter; Alonso, Juan J.: The complex-step derivative approximation (2003)
  13. Burkardt, John; Gunzburger, Max; Peterson, Janet: Insensitive functionals, inconsistent gradients, spurious minima, and regularized functionals in flow optimization problems (2002)
  14. Fagan, M.; Carle, A.: Switchback: Profile-driven recomputation for reverse mode (2002)
  15. Hovland, P.D.; McInnes, L.C.: Parallel simulation of compressible flow using automatic differentiation and PETSc (2001)
  16. Daescu, Dacian; Carmichael, Gregory R.; Sandu, Adrian: Adjoint implementation of Rosenbrock methods applied to variational data assimilation problems. (2000)
  17. Gunzburger, Max: Adjoint equation-based methods for control problems in incompressible, viscous flows (2000)
  18. Bartholomew-Biggs, M.C.: Using forward accumulation for automatic differentiation of implicitly-defined functions (1998)
  19. Glowinski, Roland; He, Jiwen: On shape optimization and related issues (1998)
  20. Gunzburger, Max: Sensitivities in computational methods for optimal flow control (1998)

1 2 next