gofee.py 18 KB
Newer Older
Malthe Kjær Bisbo's avatar
Malthe Kjær Bisbo committed
1
import numpy as np
Malthe Kjær Bisbo's avatar
update    
Malthe Kjær Bisbo committed
2
3
import pickle
from os.path import isfile
Malthe Kjær Bisbo's avatar
Malthe Kjær Bisbo committed
4

5
6
7
8
9
10
from surrogate.gpr import GPR
from population import population

from ase import Atoms
from ase.io import read, write, Trajectory
from ase.calculators.singlepoint import SinglePointCalculator
11
from ase.calculators.dftb import Dftb
12
13

from parallel_utils import split, parallel_function_eval
14

Malthe Kjær Bisbo's avatar
Malthe Kjær Bisbo committed
15
16
from bfgslinesearch_constrained import relax

17
from candidate_operations.candidate_generation import CandidateGenerator
18
19
from candidate_operations.basic_mutations import RattleMutation

20
21
22
23
24
25
26
from mpi4py import MPI
world = MPI.COMM_WORLD

import traceback
import sys

class GOFEE():
27
28
    """GOFEE global structure search method.
        
29
30
31
32
33
34
35
36
    structures: Atoms-object, list of Atoms-objects or None.
    In initial structures from which to start the sesarch.
    If None, the startgenerator must be supplied.
    If less than Ninit structures is supplied, the remaining
    ones are generated using the startgenerator or by rattling
    the supplied structures, depending on wether the
    startgenerator is supplied.

37
38
39
40
    calc: ASE calculator specifying the energy-expression
    with respect to which the atomic coordinates are
    globally optimized.

41
42
43
    gpr: The Gaussian Process Regression model used as the
    surrogate model for the Potential energy surface.
    
44
45
46
47
    startgenerator: Object used to generate initial random
    structures. Must be supplied if structures if structues=None.
    (This is the recommended way to initialize the search.)

48
49
    candidate_generator: OperationSelector object
    Object used to generate new candidates.
50

51
52
    trajectory: Trajectory object or str
    Name of trajectory to which all structures,
53
54
    evaluated during the search, is saved.

55
56
    kappa: float
    How much to weigh predicted uncertainty in acquisition
57
58
    function.

59
60
    max_steps: int
    Number of search steps.
61

62
63
    Ninit: int
    Number of initial structures. If len(structures) <
64
65
66
67
    Ninit, the remaining structures are generated using the
    startgenerator (if supplied) or by rattling the supplied
    'structures'.

68
69
    dmax_cov: float
    Max distance that an atom is allowed to move during
70
71
    surrogate relaxation (in units of covalent distance).

72
73
    Ncandidates: int
    Number of new cancidate structures generated and
74
    surrogate-relaxed in each search iteration.
75

76
77
    population_size: int
    Size of population.
78

79
80
81
82
83
84
85
86
87
88
    dualpoint: boolean
    Whether to use dualpoint evaluation or not.

    min_certainty: float
    Max predicted uncertainty allowed for structures to be
    considdered for evaluation. (in units of the maximum possible
    uncertainty.)

    restart: str
    Filename for restart file.  Default value is *None*.
89
    """
90
91
92
93
94
    def __init__(self, structures=None,
                 calc=None,
                 gpr=None,
                 startgenerator=None,
                 candidate_generator=None,
95
                 trajectory='structures.traj',
96
                 kappa=2,
Malthe Kjær Bisbo's avatar
Malthe Kjær Bisbo committed
97
                 max_steps=200,
Malthe Kjær Bisbo's avatar
update    
Malthe Kjær Bisbo committed
98
                 Ninit=10,
99
100
101
102
                 dmax_cov=3.5,
                 Ncandidates=30,
                 population_size=5,
                 dualpoint=True,
Malthe Kjær Bisbo's avatar
update    
Malthe Kjær Bisbo committed
103
                 min_certainty=0.7,
104
                 restart=None):
105
        
106
107
108
109
        if structures is None:
            assert startgenerator is not None
            self.structures = None
        else:
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
            if isinstance(structures, Atoms):
                self.structures = [structures]
            elif isinstance(structures, list):
                assert isinstance(structures[0], Atoms)
                self.structures = structures
            elif isinstance(structures, str):
                self.structures = read(structures, index=':')
        
        if calc is not None:
            self.calc = calc
        else:
            assert structures is not None
            calc = structures[0].get_calculator()
            assert calc is not None and not isinstance(calc, SinglePointCalculator)
            print('Using calculator from supplied structure(s)')
            self.calc = calc

        if gpr is not None:
            self.gpr = gpr
        else:
            self.gpr = GPR()

132
        if startgenerator is None:
133
134
            assert structures is not None
            self.startgenerator = None
135
136
        else:
            self.startgenerator = startgenerator
137

138
139
140
141
142
143
144
145
146
        if startgenerator is not None:
            self.n_to_optimize = len(self.startgenerator.stoichiometry)
        else:
            self.n_to_optimize = len(self.structures[0])
            for constraint in self.structures[0].constraints:
                if isinstance(constraint, FixAtoms):
                    indices_fixed = constraint.get_indices()
                    self.n_to_optimize -= len(indices_fixed)
                    break
Malthe Kjær Bisbo's avatar
Malthe Kjær Bisbo committed
147
        
148
149
150
        if candidate_generator is not None:
            self.candidate_generator = candidate_generator
        else:
151
152
153
            rattle = RattleMutation(self.n_to_optimize,
                                    Nrattle=3,
                                    rattle_range=4)
154
            self.candidate_generator = CandidateGenerator([1.0],[rattle])
155
156
157
158
159
160
161
162
163

        # Initialize population
        self.population = population(population_size=population_size, gpr=self.gpr, similarity2equal=0.9999)

        # Define parallel communication
        self.comm = world.Dup()  # Important to avoid mpi-problems from call to ase.parallel in BFGS
        self.master = self.comm.rank == 0

        self.kappa = kappa
Malthe Kjær Bisbo's avatar
Malthe Kjær Bisbo committed
164
        self.max_steps = max_steps
Malthe Kjær Bisbo's avatar
update    
Malthe Kjær Bisbo committed
165
        self.Ninit = Ninit
166
167
168
169
        self.dmax_cov = dmax_cov
        self.Ncandidates = Ncandidates
        self.dualpoint = dualpoint
        self.min_certainty = min_certainty
Malthe Kjær Bisbo's avatar
update    
Malthe Kjær Bisbo committed
170
171
172
173
174
175
176
177
178
179
        self.restart = restart
        
        if isinstance(trajectory, str):
            self.trajectory = Trajectory(filename=trajectory, mode='a', master=self.master)
            if self.restart:
                self.traj_name = trajectory
        elif isinstance(trajectory, Trajectory):
            self.trajectory = trajectory
        else:
            assert trajectory is None
180

Malthe Kjær Bisbo's avatar
update    
Malthe Kjær Bisbo committed
181
182
183
184
185
186
187
        if restart is None or not isfile(restart):
            self.initialize()
        else:
            self.read()
            self.comm.barrier()

    def initialize(self):
188
        self.get_initial_structures()
Malthe Kjær Bisbo's avatar
update    
Malthe Kjær Bisbo committed
189
        self.steps = 0
190
191

    def get_initial_structures(self):
192
193
194
195
196
197
198
199
200
        """Method to prepare the initial structures for the search.
        
        The method makes sure that there are atleast self.Ninit
        initial structures.
        These structures are first of all the potentially supplied
        structures. If more structures are required, these are
        generated using self.startgenerator (if supplied), otherwise
        they are generated by heavily rattling the supplied structures.
        """
Malthe Kjær Bisbo's avatar
Malthe Kjær Bisbo committed
201
        
202
        # Collect potentially supplied structures.
Malthe Kjær Bisbo's avatar
update    
Malthe Kjær Bisbo committed
203
        if self.structures is not None:
204
            for a in self.structures:
205
                a.info = {'origin': 'PreSupplied'}
206
207
        else:
            self.structures = []
Malthe Kjær Bisbo's avatar
Malthe Kjær Bisbo committed
208
        
209
        Nremaining = self.Ninit - len(self.structures)
Malthe Kjær Bisbo's avatar
Malthe Kjær Bisbo committed
210
        
211
212
213
214
215
216
217
218
219
220
221
222
        if Nremaining > 0 and self.startgenerator is None:
            # Initialize rattle-mutation for all atoms.
            rattle = RattleMutation(self.n_to_optimize,
                                    Nrattle=self.n_to_optimize,
                                    rattle_range=2)

        # Generation of remaining initial-structures (up to self.Ninit).
        for i in range(Nremaining):
            if self.startgenerator is not None:
                a = self.startgenerator.get_new_candidate()
            else:
                # Perform two times rattle of all atoms.
223
                a0 = self.structures[i % len(self.structures)]
224
225
                a = rattle.get_new_candidate([a])
                a = rattle.get_new_candidate([a])
226
227
228
229
230
231
232
233
234
235
236
            self.structures.append(a)
                
    def evaluate_initial_structures(self):
        structures_init = []
        for a in self.structures:
            calc = a.get_calculator()
            if isinstance(calc, SinglePointCalculator):
                if 'energy' in calc.results and 'forces' in calc.results:
                    # Write without evaluating.
                    structures_init.append(a)
                    continue
237
238
            a = self.evaluate(a)
            structures_init.append(a)
239

240
241
        for a in structures_init:
            self.write(a)
242
243
        self.gpr.memory.save_data(structures_init)
        self.population.add(structures_init)
244

245
    def run(self):
246
247
        """ Method to run the search.
        """
248
249
        if self.steps == 0:
            self.evaluate_initial_structures()
250

Malthe Kjær Bisbo's avatar
Malthe Kjær Bisbo committed
251
        while self.steps < self.max_steps:
252
            self.print_master('\n ### steps: {} ###\n'.format(self.steps))
253
254
255
256
257
            self.train_surrogate()
            self.update_population()
            relaxed_candidates = self.get_surrogate_relaxed_candidates()

            kappa = self.kappa
Malthe Kjær Bisbo's avatar
update    
Malthe Kjær Bisbo committed
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
            a_add = []
            for _ in range(5):
                try:
                    anew = self.select_with_acquisition(relaxed_candidates, kappa)
                    anew = self.evaluate(anew)
                    a_add.append(anew)
                    if self.dualpoint:
                        adp = self.get_dualpoint(anew)
                        adp = self.evaluate(adp)
                        a_add.append(adp)
                    break
                except Exception as err:
                    kappa /=2
                    if self.master:
                        traceback.print_exc(file=sys.stderr)
            self.gpr.memory.save_data(a_add)
274
275

            # Add structure to population
276
277
            index_lowest = np.argmin([a.get_potential_energy() for a in a_add])
            self.population.add([a_add[index_lowest]])
Malthe Kjær Bisbo's avatar
update    
Malthe Kjær Bisbo committed
278
279
280
281

            # Save search state
            self.dump((self.steps, self.population, np.random.get_state()))
            
282
283
            if self.master:
                print('anew pred:', anew.info['key_value_pairs']['Epred'], anew.info['key_value_pairs']['Epred_std'])
284
                print('E_true:', [a.get_potential_energy() for a in a_add])
285
                print('pop:', [a.get_potential_energy() for a in self.population.pop])
286
287

            self.steps += 1
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
            
    def get_dualpoint(self, a, lmax=0.10, Fmax_flat=5):
        """Returns dual-point structure, i.e. the original structure
        perturbed slightly along the forces.
        
        lmax: The atom with the largest force will be displaced by
        this distance
        
        Fmax_flat: maximum atomic displacement. is increased linearely
        with force until Fmax = Fmax_flat, after which it remains
        constant as lmax.
        """
        F = a.get_forces()
        a_dp = a.copy()

        # Calculate and set new positions
        Fmax = np.sqrt((F**2).sum(axis=1).max())
        pos_displace = lmax * F*min(1/Fmax_flat, 1/Fmax)
        pos_dp = a.positions + pos_displace
        a_dp.set_positions(pos_dp)
        return a_dp

    def print_master(self, *args):
        self.comm.barrier()
        if self.master:
            print(*args, flush=True)

    def get_surrogate_relaxed_candidates(self):
316
317
318
319
320
        """ Method supplying a number of surrogate-relaxed new
        candidates. The method combines the generation of new
        candidates with subsequent surrogate relaxation.
        The tasks are parrlelized over all avaliable cores.
        """
321
322
323
324
325
326
327
328
329
330
331
332
        Njobs = self.Ncandidates
        task_split = split(Njobs, self.comm.size)
        def func1():
            return [self.generate_candidate() for i in task_split[self.comm.rank]]
        candidates = parallel_function_eval(self.comm, func1)
        
        def func2():
            return [self.surrogate_relaxation(candidates[i], Fmax=0.1, steps=200, kappa=self.kappa)
                    for i in task_split[self.comm.rank]]
        relaxed_candidates = parallel_function_eval(self.comm, func2)
        relaxed_candidates = self.certainty_filter(relaxed_candidates)

Malthe Kjær Bisbo's avatar
Malthe Kjær Bisbo committed
333
        relaxed_candidates = self.population.pop_MLrelaxed + relaxed_candidates
334
335
336
337
338
339
340
        
        if self.master:
            Epred = np.array([a.info['key_value_pairs']['Epred'] for a in relaxed_candidates])
            Epred_std = np.array([a.info['key_value_pairs']['Epred_std'] for a in relaxed_candidates])
            fitness = Epred - self.kappa*Epred_std
            print(np.c_[Epred, Epred_std, fitness])
        return relaxed_candidates
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364

    def generate_candidate(self):
        """ Method to generate new candidate.
        """
        parents = self.population.get_structure_pair()
        a_mutated = self.candidate_generator.get_new_candidate(parents)
        return a_mutated

    def surrogate_relaxation(self, a, Fmax=0.1, steps=200, kappa=None):
        """ Method to carry out relaxations of new candidates in the
        surrogate potential.
        """
        calc = self.gpr.get_calculator(kappa)
        a_relaxed = relax(a, calc, dmax_cov=self.dmax_cov, Fmax=Fmax, steps_max=steps)

        # Evaluate uncertainty
        E, Estd = self.gpr.predict_energy(a_relaxed, eval_std=True)

        # Save prediction in info-dict
        a_relaxed.info['key_value_pairs']['Epred'] = E
        a_relaxed.info['key_value_pairs']['Epred_std'] = Estd
        a_relaxed.info['key_value_pairs']['kappa'] = self.kappa
        
        return a_relaxed
365
366
        
    def certainty_filter(self, structures):
367
368
369
370
        """ Method to filter away the most uncertain surrogate-relaxed
        candidates, which might otherewise get picked for first-principles
        evaluation, based on the very high uncertainty alone.
        """
371
372
373
374
375
376
377
378
379
380
381
382
383
        certainty = np.array([a.info['key_value_pairs']['Epred_std']
                              for a in structures]) / np.sqrt(self.gpr.K0)
        min_certainty = self.min_certainty
        for _ in range(5):
            filt = certainty < min_certainty
            if np.sum(filt.astype(int)) > 0:
                structures = [structures[i] for i in range(len(filt)) if filt[i]]
                break
            else:
                min_certainty = min_certainty + (1-min_certainty)/2
        return structures

    def update_population(self):
384
385
386
        """ Method to update the population with the new pirst-principles
        evaluated structures.
        """
387
388
389
390
391
392
393
394
395
396
397
        Njobs = len(self.population.pop)
        task_split = split(Njobs, self.comm.size)
        func = lambda: [self.surrogate_relaxation(self.population.pop[i],
                                                  Fmax=0.001, steps=200, kappa=None)
                        for i in task_split[self.comm.rank]]
        self.population.pop_MLrelaxed = parallel_function_eval(self.comm, func)
        if self.master:
            print('ML-relaxed pop forces:\n',
                  [(a.get_forces()**2).sum(axis=1).max()**0.5 for a in self.population.pop_MLrelaxed])

    def train_surrogate(self):
398
399
400
401
402
403
        """ Method to train the surrogate model.
        The method only performs hyperparameter optimization every 
        ten training instance, as carrying out the hyperparameter
        optimization is significantly more expensive than the basic
        training.
        """
404
        # Train
Malthe Kjær Bisbo's avatar
update    
Malthe Kjær Bisbo committed
405
        if self.steps < 50 or (self.steps % 10) == 0:
406
407
408
409
410
411
412
413
            self.gpr.optimize_hyperparameters(comm=self.comm)
        else:
            self.gpr.train()
        if self.master:
            print('kernel:', list(np.exp(self.gpr.kernel.theta)))
            print('lml:', self.gpr.lml)

    def select_with_acquisition(self, structures, kappa):
414
415
416
417
        """ Method to select single most "promizing" candidate 
        for first-principles evaluation according to the acquisition
        function min(E-kappa*std(E)).
        """
418
419
420
421
422
423
424
425
426
        Epred = np.array([a.info['key_value_pairs']['Epred']
                          for a in structures])
        Epred_std = np.array([a.info['key_value_pairs']['Epred_std']
                              for a in structures])
        acquisition = Epred - kappa*Epred_std
        index_select = np.argmin(acquisition)
        return structures[index_select]

    def evaluate(self, a):
427
428
429
        """ Method to evaluate the energy and forces of the selacted
        candidate.
        """
Malthe Kjær Bisbo's avatar
update    
Malthe Kjær Bisbo committed
430
        a = self.comm.bcast(a, root=0)
431
432
433
        a.wrap()

        if isinstance(self.calc, Dftb):
434
            if self.master:
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
                try:
                    a.set_calculator(self.calc)
                    E = a.get_potential_energy()
                    F = a.get_forces()
                    results = {'energy': E, 'forces': F}
                    calc_sp = SinglePointCalculator(a, **results)
                    a.set_calculator(calc_sp)
                    success = True
                except:
                    print('dftb failed on master', flush=True)
                    success = False
            else:
                success = None
            success = self.comm.bcast(success, root=0)
            if success == False:
                write('fail.traj', a)
                print('Raising error rank', self.comm.rank, flush=True)
                raise RuntimeError('DFTB evaluation failed')
453
454
455
456
457
458
459
460
            a = self.comm.bcast(a, root=0)
        else:
            a.set_calculator(self.calc)
            E = a.get_potential_energy()
            F = a.get_forces()
            results = {'energy': E, 'forces': F}
            calc_sp = SinglePointCalculator(a, **results)
            a.set_calculator(calc_sp)
461
462
463
464
465
466

        self.write(a)

        return a

    def write(self, a):
467
468
        """ Method for writing new evaluated structures to file.
        """
469
470
471
        if self.trajectory is not None:
            self.trajectory.write(a)

Malthe Kjær Bisbo's avatar
update    
Malthe Kjær Bisbo committed
472
    def dump(self, data):
473
474
475
        """ Method to save restart-file used if the search is
        restarted from some point in the search. 
        """
Malthe Kjær Bisbo's avatar
update    
Malthe Kjær Bisbo committed
476
477
        if self.comm.rank == 0 and self.restart is not None:
            pickle.dump(data, open(self.restart, "wb"), protocol=2)
478

Malthe Kjær Bisbo's avatar
update    
Malthe Kjær Bisbo committed
479
    def read(self):
480
481
482
        """ Method to restart a search from the restart-file and the
        trajectory-file containing all structures evaluated so far.
        """
Malthe Kjær Bisbo's avatar
update    
Malthe Kjær Bisbo committed
483
484
485
486
        self.steps, self.population, random_state = pickle.load(open(self.restart, "rb"))
        np.random.set_state(random_state)
        training_structures = read(self.traj_name, index=':')
        self.gpr.memory.save_data(training_structures)