source: src/Potentials/helpers.hpp@ 4464ef

Action_Thermostats Add_AtomRandomPerturbation Add_RotateAroundBondAction Add_SelectAtomByNameAction Adding_Graph_to_ChangeBondActions Adding_MD_integration_tests Adding_StructOpt_integration_tests Automaking_mpqc_open AutomationFragmentation_failures Candidate_v1.6.0 Candidate_v1.6.1 ChangeBugEmailaddress ChangingTestPorts ChemicalSpaceEvaluator Combining_Subpackages Debian_Package_split Debian_package_split_molecuildergui_only Disabling_MemDebug Docu_Python_wait EmpiricalPotential_contain_HomologyGraph_documentation Enable_parallel_make_install Enhance_userguide Enhanced_StructuralOptimization Enhanced_StructuralOptimization_continued Example_ManyWaysToTranslateAtom Exclude_Hydrogens_annealWithBondGraph FitPartialCharges_GlobalError Fix_ChronosMutex Fix_StatusMsg Fix_StepWorldTime_single_argument Fix_Verbose_Codepatterns ForceAnnealing_goodresults ForceAnnealing_oldresults ForceAnnealing_tocheck ForceAnnealing_with_BondGraph ForceAnnealing_with_BondGraph_continued ForceAnnealing_with_BondGraph_continued_betteresults ForceAnnealing_with_BondGraph_contraction-expansion GeometryObjects Gui_displays_atomic_force_velocity IndependentFragmentGrids_IntegrationTest JobMarket_RobustOnKillsSegFaults JobMarket_StableWorkerPool JobMarket_unresolvable_hostname_fix ODR_violation_mpqc_open PartialCharges_OrthogonalSummation PythonUI_with_named_parameters QtGui_reactivate_TimeChanged_changes Recreated_GuiChecks RotateToPrincipalAxisSystem_UndoRedo StoppableMakroAction Subpackage_levmar Subpackage_vmg ThirdParty_MPQC_rebuilt_buildsystem TremoloParser_IncreasedPrecision TremoloParser_MultipleTimesteps Ubuntu_1604_changes stable
Last change on this file since 4464ef was e1fe7e, checked in by Frederik Heber <heber@…>, 11 years ago

FunctionModel now uses list_of_arguments to split sequence of subsets of distances.

  • this fixes ambiguities with the set of distances: Imagine the distances within a water molecule as OH (A) and HH (B). We then may have a sequence of argument_t as AABAAB. And with the current implementation of CompoundPotential::splitUpArgumentsByModels() we would always choose the latter (and more complex) model. Hence, we make two calls to TriplePotential_Angle, instead of calls twice to PairPotential_Harmonic for A, one to PairPotential_Harmonic for B, and once to TriplePotential_Angle for AAB.
  • now, we new list looks like A,A,B,AAB where each tuple of distances can be uniquely associated with a specific potential.
  • changed signatures of EmpiricalPotential::operator(), ::derivative(), ::parameter_derivative(). This involved changing all of the current specific potentials and CompoundPotential.
  • TrainingData must discern between the InputVector_t (just all distances) and the FilteredInputVector_t (tuples of subsets of distances).
  • FunctionApproximation now has list_of_arguments_t as parameter to evaluate() and evaluate_derivative().
  • DOCU: docu change in TrainingData.
  • Property mode set to 100644
File size: 2.6 KB
RevLine 
[6bb72a]1/*
2 * helpers.hpp
3 *
4 * Created on: Sep 26, 2012
5 * Author: heber
6 */
7
8#ifndef POTENTIALS_HELPERS_HPP_
9#define POTENTIALS_HELPERS_HPP_
10
11
12// include config.h
13#ifdef HAVE_CONFIG_H
14#include <config.h>
15#endif
16
[dda3e8]17#include <cmath>
18//#include <iomanip>
19#include <limits>
20//#include <sstream>
21
[775dd1a]22#include "FunctionApproximation/FunctionModel.hpp"
23#include "FunctionApproximation/FunctionArgument.hpp"
24
[dda3e8]25//#include "CodePatterns/Log.hpp"
26
[6bb72a]27namespace Helpers
28{
29
30 /** Integer-optimized version for power of n.
31 *
32 * This is taken from Julian Iseringhausen's VMG project which is under
33 * GPL.
34 *
35 * @param base base
36 * @param power power to take \a base to
37 * @return \a base to the power of \a power
38 */
39 inline int intpow(int base, unsigned int power)
40 {
41 int result = 1;
42 while (power != 0) {
43 if (power & 1)
44 result *= base;
45 base *= base;
46 power >>= 1;
47 }
48 return result;
49 }
50
51 /** Double-optimized version for power of n.
52 *
53 * This is taken from Julian Iseringhausen's VMG project which is under
54 * GPL.
55 *
56 * @param base base
57 * @param power power to take \a base to
58 * @return \a base to the power of \a power
59 */
60 inline double pow(double base, unsigned int power)
61 {
62 double result = 1.0;
63 while (power != 0) {
64 if (power & 1)
65 result *= base;
66 base *= base;
67 power >>= 1;
68 }
69 return result;
70 }
71
[dda3e8]72 /** Equality operator that takes numerical precision into account.
73 *
74 * \param first first operand
75 * \param second second operand
76 * \param factor factor to influence numeric threshold
77 */
78 template <typename T>
79 inline bool isEqual(const T &first, const T &second, const double factor=1.) {
80// std::stringstream stream;
81// stream << std::setprecision(10)
82// << "Comparing " << first << " to " << second << ": "
83// << fabs(first-second) << "<" << std::numeric_limits<T>::epsilon()*factor << "?";
84// LOG(1, stream.str());
85 return (fabs(first-second) < std::numeric_limits<T>::epsilon()*factor);
86 }
87
[775dd1a]88 inline std::vector< FunctionModel::arguments_t >
89 NoOp_Triplefunction(
90 const argument_t &,
91 const double)
92 {
93 return std::vector< FunctionModel::arguments_t >();
94 }
95
[0f5d38]96 inline FunctionModel::arguments_t
97 NoOp_Filterfunction(
98 const FunctionModel::arguments_t &args
99 )
100 {
101 return args;
102 }
103
104 inline FunctionModel::arguments_t
105 returnEmptyArguments()
106 {
107 return FunctionModel::arguments_t();
108 }
109
[e1fe7e]110 inline FunctionModel::list_of_arguments_t
111 returnEmptyListArguments()
112 {
113 return FunctionModel::list_of_arguments_t();
114 }
115
[6bb72a]116}; /* namespace Helpers */
117
118
119#endif /* POTENTIALS_HELPERS_HPP_ */
Note: See TracBrowser for help on using the repository browser.