Semiautomatic differentiation for efficient gradient computations Many large-scale computations involve a mesh and first (or sometimes higher) partial derivatives of functions of mesh elements. In principle, automatic differentiation (AD) can provide the requisite partials more efficiently and accurately than conventional finite-difference approximations. AD requires source-code modifications, which may be little more than changes to declarations. Such simple changes can easily give improved results, e.g., when Jacobian-vector products are used iteratively to solve nonlinear equations. When gradients are required (say, for optimization) and the problem involves many variables, “backward AD” in theory is very efficient, but when carried out automatically and straightforwardly, may use a prohibitive amount of memory. In this case, applying AD separately to each element function and manually assembling the gradient pieces -- semiautomatic differentiation -- can deliver gradients efficiently and accurately. This paper concerns on-going work; it compares several implementations of backward AD, describes a simple operator-overloading implementation specialized for gradient computations, and compares the implementations on some mesh-optimization examples. Ideas from the specialized implementation could be used in fully general source-to-source translators for C and C++.
Keywords for this software
References in zbMATH (referenced in 9 articles )
Showing results 1 to 9 of 9.
- Brothers, Michael D.; Foster, John T.; Millwater, Harry R.: A comparison of different methods for calculating tangent-stiffness matrices in a massively parallel computational peridynamics code (2014)
- Hogan, Robin J.: Fast reverse-mode automatic differentiation using expression templates in C++ (2014)
- Papamarkou, Theodore; Mira, Antonietta; Girolami, Mark: Zero variance differential geometric Markov chain Monte Carlo algorithms (2014)
- Bücker, H. Martin; Slusanschi, Emil: Second-order derivatives of the general-purpose finite element package SEPRAN via source transformation (2011)
- Enciu, P.; Wurtz, F.; Gerbaud, L.; Delinchant, B.: Automatic differentiation for electromagnetic models used in optimization (2009)
- Bücker, H. Martin; Petera, Monika; Vehreschild, Andre: Code optimization techniques in source transformations for interpreted languages (2008)
- Phipps, Eric T.; Bartlett, Roscoe A.; Gay, David M.; Hoekstra, Robert J.: Large-scale transient sensitivity analysis of a radiation-damaged bipolar junction transistor via automatic differentiation (2008)
- Bartlett, Roscoe A.; Gay, David M.; Phipps, Eric T.: Automatic differentiation of C++ codes for large-scale scientific computing (2006)
- Gay, David M.: Semiautomatic differentiation for efficient gradient computations (2006)