Changeset 715
- Timestamp:
- 04/03/14 15:37:12 (2 years ago)
- Location:
- mystic
- Files:
-
- 3 edited
- 2 moved
Legend:
- Unmodified
- Added
- Removed
-
mystic/README.md
r701 r715 55 55 http://mmckerns.github.io/project/mystic 56 56 57 Mystic is distributed under a modifiedBSD license.57 Mystic is distributed under a 3-clause BSD license. 58 58 59 59 Development Release -
mystic/mystic/abstract_ensemble_solver.py
r713 r715 6 6 # - http://mmckerns.github.io/project/mystic/browser/mystic/LICENSE 7 7 # 8 # Abstract NestedSolver Class8 # Abstract Ensemble Solver Class 9 9 """ 10 10 This module contains the base class for launching several mystic solvers 11 11 instances -- utilizing a parallel "map" function to enable parallel 12 computing. This module describes the nestedsolver interface. As with12 computing. This module describes the ensemble solver interface. As with 13 13 the AbstractSolver, the "Solve" method must be overwritten with the derived 14 14 solver's optimization algorithm. Similar to AbstractMapSolver, a call to … … 26 26 ===== 27 27 28 A typical call to a ' nested' solver will roughly follow this example:28 A typical call to a 'ensemble' solver will roughly follow this example: 29 29 30 30 >>> # the function to be minimized and the initial values … … 75 75 76 76 """ 77 __all__ = ['Abstract NestedSolver']77 __all__ = ['AbstractEnsembleSolver'] 78 78 79 79 … … 82 82 83 83 84 class Abstract NestedSolver(AbstractMapSolver):84 class AbstractEnsembleSolver(AbstractMapSolver): 85 85 """ 86 Abstract NestedSolver base class for mystic optimizers that are nested within86 AbstractEnsembleSolver base class for mystic optimizers that are called within 87 87 a parallel map. This allows pseudo-global coverage of parameter space using 88 88 non-global optimizers. … … 111 111 signal_handler - catches the interrupt signal. [***disabled***] 112 112 """ 113 super(Abstract NestedSolver, self).__init__(dim, **kwds)113 super(AbstractEnsembleSolver, self).__init__(dim, **kwds) 114 114 #self.signal_handler = None 115 115 #self._handle_sigint = False -
mystic/mystic/ensemble.py
r713 r715 9 9 ======= 10 10 11 This module contains a collection of optimization that use map-reduce11 This module contains a collection of optimization routines that use "map" 12 12 to distribute several optimizer instances over parameter space. Each 13 13 solver accepts a imported solver object as the "nested" solver, which 14 14 becomes the target of the map function. 15 15 16 The set of solvers built on mystic's Abstract NestdSolver are::16 The set of solvers built on mystic's AbstractEnsembleSolver are:: 17 17 LatticeSolver -- start from center of N grid points 18 18 BuckshotSolver -- start from N random points in parameter space … … 33 33 from mystic.tools import wrap_function 34 34 35 from mystic.abstract_ nested_solver import AbstractNestedSolver36 37 38 class LatticeSolver(Abstract NestedSolver):35 from mystic.abstract_ensemble_solver import AbstractEnsembleSolver 36 37 38 class LatticeSolver(AbstractEnsembleSolver): 39 39 """ 40 40 parallel mapped optimization starting from the center of N grid points … … 46 46 nbins -- tuple of number of bins in each dimension 47 47 48 All important class members are inherited from Abstract NestedSolver.48 All important class members are inherited from AbstractEnsembleSolver. 49 49 """ 50 50 super(LatticeSolver, self).__init__(dim, nbins=nbins) … … 111 111 112 112 # get the nested solver instance 113 solver = self._Abstract NestedSolver__get_solver_instance()113 solver = self._AbstractEnsembleSolver__get_solver_instance() 114 114 #------------------------------------------------------------- 115 115 … … 204 204 return 205 205 206 class BuckshotSolver(Abstract NestedSolver):206 class BuckshotSolver(AbstractEnsembleSolver): 207 207 """ 208 208 parallel mapped optimization starting from the N random points … … 214 214 npts -- number of parallel solver instances 215 215 216 All important class members are inherited from Abstract NestedSolver.216 All important class members are inherited from AbstractEnsembleSolver. 217 217 """ 218 218 super(BuckshotSolver, self).__init__(dim, npts=npts) … … 275 275 276 276 # get the nested solver instance 277 solver = self._Abstract NestedSolver__get_solver_instance()277 solver = self._AbstractEnsembleSolver__get_solver_instance() 278 278 #------------------------------------------------------------- 279 279 -
mystic/mystic/solvers.py
r713 r715 47 47 48 48 For more information, please see the solver documentation found here:: 49 - mystic.mystic.differential_evolution [differential evolution solvers]50 - mystic.mystic.scipy_optimize [scipy local-search solvers]51 - mystic.mystic. nested[pseudo-global solvers]49 - mystic.mystic.differential_evolution [differential evolution solvers] 50 - mystic.mystic.scipy_optimize [scipy local-search solvers] 51 - mystic.mystic.ensemble [pseudo-global solvers] 52 52 53 53 or the API documentation found here:: 54 - mystic.mystic.abstract_solver [the solver API definition]55 - mystic.mystic.abstract_map_solver [the parallel solver API]56 - mystic.mystic.abstract_ nested_solver [the nestedsolver API]54 - mystic.mystic.abstract_solver [the solver API definition] 55 - mystic.mystic.abstract_map_solver [the parallel solver API] 56 - mystic.mystic.abstract_ensemble_solver [the ensemble solver API] 57 57 58 58 … … 64 64 65 65 # pseudo-global optimizers 66 from nestedimport BuckshotSolver67 from nestedimport LatticeSolver66 from ensemble import BuckshotSolver 67 from ensemble import LatticeSolver 68 68 69 69 # local-search optimizers -
mystic/setup.py
r713 r715 203 203 by the `pathos` package, and thus with little new code solvers are 204 204 extended to high-performance computing. For more information, see 205 `mystic.mystic.abstract_map_solver`, `mystic.mystic.abstract_ nested_solver`,205 `mystic.mystic.abstract_map_solver`, `mystic.mystic.abstract_ensemble_solver`, 206 206 and the pathos documentation at http://dev.danse.us/trac/pathos. 207 207 208 208 Important classes and functions are found here:: 209 209 210 - mystic.mystic.solvers [solver optimization algorithms]211 - mystic.mystic.termination [solver termination conditions]212 - mystic.mystic.strategy [solver population mutation strategies]213 - mystic.mystic.monitors [optimization monitors]214 - mystic.mystic.tools [function wrappers, etc]215 - mystic.mystic.forward_model [cost function generator]216 - mystic.models [a collection of standard models]217 - mystic.math [some mathematical functions and tools]210 - mystic.mystic.solvers [solver optimization algorithms] 211 - mystic.mystic.termination [solver termination conditions] 212 - mystic.mystic.strategy [solver population mutation strategies] 213 - mystic.mystic.monitors [optimization monitors] 214 - mystic.mystic.tools [function wrappers, etc] 215 - mystic.mystic.forward_model [cost function generator] 216 - mystic.models [a collection of standard models] 217 - mystic.math [some mathematical functions and tools] 218 218 219 219 Solver and model API definitions are found here:: 220 220 221 - mystic.mystic.abstract_solver [the solver API definition]222 - mystic.mystic.abstract_map_solver [the parallel solver API]223 - mystic.mystic.abstract_ nested_solver [the nestedsolver API]224 - mystic.models.abstract_model [the model API definition]221 - mystic.mystic.abstract_solver [the solver API definition] 222 - mystic.mystic.abstract_map_solver [the parallel solver API] 223 - mystic.mystic.abstract_ensemble_solver [the ensemble solver API] 224 - mystic.models.abstract_model [the model API definition] 225 225 226 226 … … 228 228 ======= 229 229 230 Mystic is distributed under a modifiedBSD license.230 Mystic is distributed under a 3-clause BSD license. 231 231 232 232 >>> import mystic
Note: See TracChangeset
for help on using the changeset viewer.