mcmc_based_searches.py 77.8 KB
Newer Older
Gregory Ashton's avatar
Gregory Ashton committed
1
2
""" Searches using MCMC-based methods """

3
import sys
Gregory Ashton's avatar
Gregory Ashton committed
4
import os
5
import copy
Gregory Ashton's avatar
Gregory Ashton committed
6
import logging
7
from collections import OrderedDict
8
9
10
11
12
13
14
15

import numpy as np
import matplotlib
import matplotlib.pyplot as plt
import emcee
import corner
import dill as pickle

16
17
from core import BaseSearchClass, ComputeFstat, SemiCoherentSearch
from optimal_setup_functions import get_V_estimate
Gregory Ashton's avatar
Gregory Ashton committed
18
19
20
from core import tqdm, args, earth_ephem, sun_ephem
from optimal_setup_functions import get_optimal_setup
import helper_functions
21
22


Gregory Ashton's avatar
Gregory Ashton committed
23
24
class MCMCSearch(BaseSearchClass):
    """ MCMC search using ComputeFstat"""
Gregory Ashton's avatar
Gregory Ashton committed
25
    @helper_functions.initializer
Gregory Ashton's avatar
Gregory Ashton committed
26
27
    def __init__(self, label, outdir, theta_prior, tref, minStartTime,
                 maxStartTime, sftfilepath=None, nsteps=[100, 100],
28
29
                 nwalkers=100, ntemps=1, log10temperature_min=-5,
                 theta_initial=None, scatter_val=1e-10,
30
                 binary=False, BSGL=False, minCoverFreq=None,
31
                 maxCoverFreq=None, detectors=None, earth_ephem=None,
32
                 sun_ephem=None, injectSources=None, assumeSqrtSX=None):
33
34
35
36
        """
        Parameters
        label, outdir: str
            A label and directory to read/write data from/to
37
38
        sftfilepath: str
            File patern to match SFTs
39
        theta_prior: dict
40
41
42
43
            Dictionary of priors and fixed values for the search parameters.
            For each parameters (key of the dict), if it is to be held fixed
            the value should be the constant float, if it is be searched, the
            value should be a dictionary of the prior.
44
45
46
47
        theta_initial: dict, array, (None)
            Either a dictionary of distribution about which to distribute the
            initial walkers about, an array (from which the walkers will be
            scattered by scatter_val, or  None in which case the prior is used.
48
        tref, minStartTime, maxStartTime: int
49
50
51
52
53
54
            GPS seconds of the reference time, start time and end time
        nsteps: list (m,)
            List specifying the number of steps to take, the last two entries
            give the nburn and nprod of the 'production' run, all entries
            before are for iterative initialisation steps (usually just one)
            e.g. [1000, 1000, 500].
55
56
57
58
59
60
61
62
        nwalkers, ntemps: int,
            The number of walkers and temperates to use in the parallel
            tempered PTSampler.
        log10temperature_min float < 0
            The  log_10(tmin) value, the set of betas passed to PTSampler are
            generated from np.logspace(0, log10temperature_min, ntemps).
        binary: Bool
            If true, search over binary parameters
63
        detectors: str
64
65
            Two character reference to the data to use, specify None for no
            contraint.
66
67
68
69
70
71
72
73
74
75
        minCoverFreq, maxCoverFreq: float
            Minimum and maximum instantaneous frequency which will be covered
            over the SFT time span as passed to CreateFstatInput
        earth_ephem, sun_ephem: str
            Paths of the two files containing positions of Earth and Sun,
            respectively at evenly spaced times, as passed to CreateFstatInput
            If None defaults defined in BaseSearchClass will be used

        """

Gregory Ashton's avatar
Gregory Ashton committed
76
77
        if os.path.isdir(outdir) is False:
            os.mkdir(outdir)
78
        self.add_log_file()
Gregory Ashton's avatar
Gregory Ashton committed
79
80
        logging.info(
            'Set-up MCMC search for model {} on data {}'.format(
81
                self.label, self.sftfilepath))
82
83
84
        self.pickle_path = '{}/{}_saved_data.p'.format(self.outdir, self.label)
        self.unpack_input_theta()
        self.ndim = len(self.theta_keys)
85
86
87
88
        if self.log10temperature_min:
            self.betas = np.logspace(0, self.log10temperature_min, self.ntemps)
        else:
            self.betas = None
89

90
91
92
93
94
95
96
97
        if earth_ephem is None:
            self.earth_ephem = self.earth_ephem_default
        if sun_ephem is None:
            self.sun_ephem = self.sun_ephem_default

        if args.clean and os.path.isfile(self.pickle_path):
            os.rename(self.pickle_path, self.pickle_path+".old")

98
99
100
        self.log_input()

    def log_input(self):
101
        logging.info('theta_prior = {}'.format(self.theta_prior))
102
        logging.info('nwalkers={}'.format(self.nwalkers))
103
104
105
106
        logging.info('scatter_val = {}'.format(self.scatter_val))
        logging.info('nsteps = {}'.format(self.nsteps))
        logging.info('ntemps = {}'.format(self.ntemps))
        logging.info('log10temperature_min = {}'.format(
107
            self.log10temperature_min))
108
109
110

    def inititate_search_object(self):
        logging.info('Setting up search object')
Gregory Ashton's avatar
Gregory Ashton committed
111
        self.search = ComputeFstat(
112
113
114
            tref=self.tref, sftfilepath=self.sftfilepath,
            minCoverFreq=self.minCoverFreq, maxCoverFreq=self.maxCoverFreq,
            earth_ephem=self.earth_ephem, sun_ephem=self.sun_ephem,
115
            detectors=self.detectors, BSGL=self.BSGL, transient=False,
116
            minStartTime=self.minStartTime, maxStartTime=self.maxStartTime,
117
118
            binary=self.binary, injectSources=self.injectSources,
            assumeSqrtSX=self.assumeSqrtSX)
119
120

    def logp(self, theta_vals, theta_prior, theta_keys, search):
Gregory Ashton's avatar
Gregory Ashton committed
121
        H = [self.generic_lnprior(**theta_prior[key])(p) for p, key in
122
123
124
125
126
127
             zip(theta_vals, theta_keys)]
        return np.sum(H)

    def logl(self, theta, search):
        for j, theta_i in enumerate(self.theta_idxs):
            self.fixed_theta[theta_i] = theta[j]
128
129
        FS = search.compute_fullycoherent_det_stat_single_point(
            *self.fixed_theta)
130
131
132
        return FS

    def unpack_input_theta(self):
133
        full_theta_keys = ['F0', 'F1', 'F2', 'Alpha', 'Delta']
134
135
136
        if self.binary:
            full_theta_keys += [
                'asini', 'period', 'ecc', 'tp', 'argp']
137
138
        full_theta_keys_copy = copy.copy(full_theta_keys)

139
140
        full_theta_symbols = ['$f$', '$\dot{f}$', '$\ddot{f}$', r'$\alpha$',
                              r'$\delta$']
141
142
143
144
        if self.binary:
            full_theta_symbols += [
                'asini', 'period', 'period', 'ecc', 'tp', 'argp']

145
146
        self.theta_keys = []
        fixed_theta_dict = {}
147
        for key, val in self.theta_prior.iteritems():
148
149
            if type(val) is dict:
                fixed_theta_dict[key] = 0
Gregory Ashton's avatar
Gregory Ashton committed
150
                self.theta_keys.append(key)
151
152
153
154
155
156
            elif type(val) in [float, int, np.float64]:
                fixed_theta_dict[key] = val
            else:
                raise ValueError(
                    'Type {} of {} in theta not recognised'.format(
                        type(val), key))
Gregory Ashton's avatar
Gregory Ashton committed
157
            full_theta_keys_copy.pop(full_theta_keys_copy.index(key))
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173

        if len(full_theta_keys_copy) > 0:
            raise ValueError(('Input dictionary `theta` is missing the'
                              'following keys: {}').format(
                                  full_theta_keys_copy))

        self.fixed_theta = [fixed_theta_dict[key] for key in full_theta_keys]
        self.theta_idxs = [full_theta_keys.index(k) for k in self.theta_keys]
        self.theta_symbols = [full_theta_symbols[i] for i in self.theta_idxs]

        idxs = np.argsort(self.theta_idxs)
        self.theta_idxs = [self.theta_idxs[i] for i in idxs]
        self.theta_symbols = [self.theta_symbols[i] for i in idxs]
        self.theta_keys = [self.theta_keys[i] for i in idxs]

    def check_initial_points(self, p0):
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
        for nt in range(self.ntemps):
            logging.info('Checking temperature {} chains'.format(nt))
            initial_priors = np.array([
                self.logp(p, self.theta_prior, self.theta_keys, self.search)
                for p in p0[nt]])
            number_of_initial_out_of_bounds = sum(initial_priors == -np.inf)

            if number_of_initial_out_of_bounds > 0:
                logging.warning(
                    'Of {} initial values, {} are -np.inf due to the prior'
                    .format(len(initial_priors),
                            number_of_initial_out_of_bounds))

                p0 = self.generate_new_p0_to_fix_initial_points(
                    p0, nt, initial_priors)

    def generate_new_p0_to_fix_initial_points(self, p0, nt, initial_priors):
        logging.info('Attempting to correct intial values')
        idxs = np.arange(self.nwalkers)[initial_priors == -np.inf]
        count = 0
        while sum(initial_priors == -np.inf) > 0 and count < 100:
            for j in idxs:
                p0[nt][j] = (p0[nt][np.random.randint(0, self.nwalkers)]*(
                             1+np.random.normal(0, 1e-10, self.ndim)))
            initial_priors = np.array([
                self.logp(p, self.theta_prior, self.theta_keys,
                          self.search)
                for p in p0[nt]])
            count += 1

        if sum(initial_priors == -np.inf) > 0:
            logging.info('Failed to fix initial priors')
        else:
            logging.info('Suceeded to fix initial priors')

        return p0
210

211
    def OLD_run_sampler_with_progress_bar(self, sampler, ns, p0):
212
213
        for result in tqdm(sampler.sample(p0, iterations=ns), total=ns):
            pass
Gregory Ashton's avatar
Gregory Ashton committed
214
215
        return sampler

216
217
    def setup_convergence_testing(
            self, convergence_period=10, convergence_length=10,
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
            convergence_burnin_fraction=0.25, convergence_threshold_number=10,
            convergence_threshold=1.2, convergence_prod_threshold=2):
        """
        If called, convergence testing is used during the MCMC simulation

        This uses the Gelmanr-Rubin statistic based on the ratio of between and
        within walkers variance. The original statistic was developed for
        multiple (independent) MCMC simulations, in this context we simply use
        the walkers

        Parameters
        ----------
        convergence_period: int
            period (in number of steps) at which to test convergence
        convergence_length: int
            number of steps to use in testing convergence - this should be
            large enough to measure the variance, but if it is too long
            this will result in incorect early convergence tests
        convergence_burnin_fraction: float [0, 1]
            the fraction of the burn-in period after which to start testing
        convergence_threshold_number: int
            the number of consecutive times where the test passes after which
            to break the burn-in and go to production
        convergence_threshold: float
            the threshold to use in diagnosing convergence. Gelman & Rubin
            recomend a value of 1.2, 1.1 for strict convergence
        convergence_prod_threshold: float
            the threshold to test the production values with
        """
247
248
249
250
251
252
253

        if convergence_length > convergence_period:
            raise ValueError('convergence_length must be < convergence_period')
        logging.info('Setting up convergence testing')
        self.convergence_length = convergence_length
        self.convergence_period = convergence_period
        self.convergence_burnin_fraction = convergence_burnin_fraction
254
        self.convergence_prod_threshold = convergence_prod_threshold
255
256
257
258
259
260
        self.convergence_diagnostic = []
        self.convergence_diagnosticx = []
        self.convergence_threshold_number = convergence_threshold_number
        self.convergence_threshold = convergence_threshold
        self.convergence_number = 0

261
    def get_convergence_statistic(self, i, sampler):
262
263
264
265
266
267
268
269
270
271
272
273
        s = sampler.chain[0, :, i-self.convergence_length+1:i+1, :]
        within_std = np.mean(np.var(s, axis=1), axis=0)
        per_walker_mean = np.mean(s, axis=1)
        mean = np.mean(per_walker_mean, axis=0)
        between_std = np.sqrt(np.mean((per_walker_mean-mean)**2, axis=0))
        W = within_std
        B_over_n = between_std**2 / self.convergence_period
        Vhat = ((self.convergence_period-1.)/self.convergence_period * W
                + B_over_n + B_over_n / float(self.nwalkers))
        c = Vhat/W
        self.convergence_diagnostic.append(c)
        self.convergence_diagnosticx.append(i - self.convergence_period/2)
274
275
276
277
278
279
280
281
        return c

    def convergence_test(self, i, sampler, nburn):
        if i < self.convergence_burnin_fraction*nburn:
            return False
        if np.mod(i+1, self.convergence_period) == 0:
            return False
        c = self.get_convergence_statistic(i, sampler)
282
283
        if np.all(c < self.convergence_threshold):
            self.convergence_number += 1
284
285
        else:
            self.convergence_number = 0
286
287
        return self.convergence_number > self.convergence_threshold_number

288
289
290
291
292
293
294
295
296
    def check_production_convergence(self, k):
        bools = np.any(
            np.array(self.convergence_diagnostic)[k:, :]
            > self.convergence_prod_threshold, axis=1)
        if np.any(bools):
            logging.warning(
                '{} convergence tests in the production run of {} failed'
                .format(np.sum(bools), len(bools)))

297
298
    def run_sampler(self, sampler, p0, nprod=0, nburn=0):
        if hasattr(self, 'convergence_period'):
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
            converged = False
            logging.info('Running {} burn-in steps with convergence testing'
                         .format(nburn))
            iterator = tqdm(sampler.sample(p0, iterations=nburn), total=nburn)
            for i, output in enumerate(iterator):
                if converged:
                    logging.info(
                        'Converged at {} before max number {} of steps reached'
                        .format(i, nburn))
                    self.convergence_idx = i
                    break
                else:
                    converged = self.convergence_test(i, sampler, nburn)
            iterator.close()
            logging.info('Running {} production steps'.format(nprod))
            j = nburn
            k = len(self.convergence_diagnostic)
            for result in tqdm(sampler.sample(output[0], iterations=nprod),
                               total=nprod):
                self.get_convergence_statistic(j, sampler)
                j += 1
            self.check_production_convergence(k)
321
322
323
324
325
326
            return sampler
        else:
            for result in tqdm(sampler.sample(p0, iterations=nburn+nprod),
                               total=nburn+nprod):
                pass
            return sampler
327

328
    def run(self, proposal_scale_factor=2, create_plots=True, **kwargs):
329

Gregory Ashton's avatar
Gregory Ashton committed
330
        self.old_data_is_okay_to_use = self.check_old_data_is_okay_to_use()
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
        if self.old_data_is_okay_to_use is True:
            logging.warning('Using saved data from {}'.format(
                self.pickle_path))
            d = self.get_saved_data()
            self.sampler = d['sampler']
            self.samples = d['samples']
            self.lnprobs = d['lnprobs']
            self.lnlikes = d['lnlikes']
            return

        self.inititate_search_object()

        sampler = emcee.PTSampler(
            self.ntemps, self.nwalkers, self.ndim, self.logl, self.logp,
            logpargs=(self.theta_prior, self.theta_keys, self.search),
346
            loglargs=(self.search,), betas=self.betas, a=proposal_scale_factor)
347

Gregory Ashton's avatar
Gregory Ashton committed
348
349
        p0 = self.generate_initial_p0()
        p0 = self.apply_corrections_to_p0(p0)
350
351
352
353
354
        self.check_initial_points(p0)

        ninit_steps = len(self.nsteps) - 2
        for j, n in enumerate(self.nsteps[:-2]):
            logging.info('Running {}/{} initialisation with {} steps'.format(
Gregory Ashton's avatar
Gregory Ashton committed
355
                j, ninit_steps, n))
356
            sampler = self.run_sampler(sampler, p0, nburn=n)
357
358
            logging.info("Mean acceptance fraction: {}"
                         .format(np.mean(sampler.acceptance_fraction, axis=1)))
359
360
361
            if self.ntemps > 1:
                logging.info("Tswap acceptance fraction: {}"
                             .format(sampler.tswap_acceptance_fraction))
362
363
364
365
366
367
368
            if create_plots:
                fig, axes = self.plot_walkers(sampler,
                                              symbols=self.theta_symbols,
                                              **kwargs)
                fig.tight_layout()
                fig.savefig('{}/{}_init_{}_walkers.png'.format(
                    self.outdir, self.label, j), dpi=200)
369

370
            p0 = self.get_new_p0(sampler)
Gregory Ashton's avatar
Gregory Ashton committed
371
            p0 = self.apply_corrections_to_p0(p0)
372
373
374
            self.check_initial_points(p0)
            sampler.reset()

Gregory Ashton's avatar
Gregory Ashton committed
375
376
377
378
        if len(self.nsteps) > 1:
            nburn = self.nsteps[-2]
        else:
            nburn = 0
379
380
381
        nprod = self.nsteps[-1]
        logging.info('Running final burn and prod with {} steps'.format(
            nburn+nprod))
382
        sampler = self.run_sampler(sampler, p0, nburn=nburn, nprod=nprod)
383
384
        logging.info("Mean acceptance fraction: {}"
                     .format(np.mean(sampler.acceptance_fraction, axis=1)))
385
386
387
        if self.ntemps > 1:
            logging.info("Tswap acceptance fraction: {}"
                         .format(sampler.tswap_acceptance_fraction))
388

389
390
        if create_plots:
            fig, axes = self.plot_walkers(sampler, symbols=self.theta_symbols,
391
                                          nprod=nprod, **kwargs)
392
393
394
            fig.tight_layout()
            fig.savefig('{}/{}_walkers.png'.format(self.outdir, self.label),
                        dpi=200)
395
396
397
398
399
400
401
402
403
404

        samples = sampler.chain[0, :, nburn:, :].reshape((-1, self.ndim))
        lnprobs = sampler.lnprobability[0, :, nburn:].reshape((-1))
        lnlikes = sampler.lnlikelihood[0, :, nburn:].reshape((-1))
        self.sampler = sampler
        self.samples = samples
        self.lnprobs = lnprobs
        self.lnlikes = lnlikes
        self.save_data(sampler, samples, lnprobs, lnlikes)

405
    def plot_corner(self, figsize=(7, 7),  tglitch_ratio=False,
406
407
408
                    add_prior=False, nstds=None, label_offset=0.4,
                    dpi=300, rc_context={}, **kwargs):

Gregory Ashton's avatar
Gregory Ashton committed
409
410
411
412
413
414
415
416
417
418
        if self.ndim < 2:
            with plt.rc_context(rc_context):
                fig, ax = plt.subplots(figsize=figsize)
                ax.hist(self.samples, bins=50, histtype='stepfilled')
                ax.set_xlabel(self.theta_symbols[0])

            fig.savefig('{}/{}_corner.png'.format(
                self.outdir, self.label), dpi=dpi)
            return

419
420
421
422
423
424
        with plt.rc_context(rc_context):
            fig, axes = plt.subplots(self.ndim, self.ndim,
                                     figsize=figsize)

            samples_plt = copy.copy(self.samples)
            theta_symbols_plt = copy.copy(self.theta_symbols)
425
426
            theta_symbols_plt = [s.replace('_{glitch}', r'_\textrm{glitch}')
                                 for s in theta_symbols_plt]
427
428
429
430
431

            if tglitch_ratio:
                for j, k in enumerate(self.theta_keys):
                    if k == 'tglitch':
                        s = samples_plt[:, j]
432
433
434
                        samples_plt[:, j] = (
                            s - self.minStartTime)/(
                                self.maxStartTime - self.minStartTime)
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
                        theta_symbols_plt[j] = r'$R_{\textrm{glitch}}$'

            if type(nstds) is int and 'range' not in kwargs:
                _range = []
                for j, s in enumerate(samples_plt.T):
                    median = np.median(s)
                    std = np.std(s)
                    _range.append((median - nstds*std, median + nstds*std))
            else:
                _range = None

            fig_triangle = corner.corner(samples_plt,
                                         labels=theta_symbols_plt,
                                         fig=fig,
                                         bins=50,
                                         max_n_ticks=4,
                                         plot_contours=True,
                                         plot_datapoints=True,
                                         label_kwargs={'fontsize': 8},
                                         data_kwargs={'alpha': 0.1,
                                                      'ms': 0.5},
                                         range=_range,
                                         **kwargs)

            axes_list = fig_triangle.get_axes()
            axes = np.array(axes_list).reshape(self.ndim, self.ndim)
            plt.draw()
            for ax in axes[:, 0]:
                ax.yaxis.set_label_coords(-label_offset, 0.5)
            for ax in axes[-1, :]:
                ax.xaxis.set_label_coords(0.5, -label_offset)
            for ax in axes_list:
                ax.set_rasterized(True)
                ax.set_rasterization_zorder(-10)
            plt.tight_layout(h_pad=0.0, w_pad=0.0)
            fig.subplots_adjust(hspace=0.05, wspace=0.05)

            if add_prior:
                self.add_prior_to_corner(axes, samples_plt)

            fig_triangle.savefig('{}/{}_corner.png'.format(
                self.outdir, self.label), dpi=dpi)
477
478
479
480
481
482

    def add_prior_to_corner(self, axes, samples):
        for i, key in enumerate(self.theta_keys):
            ax = axes[i][i]
            xlim = ax.get_xlim()
            s = samples[:, i]
Gregory Ashton's avatar
Gregory Ashton committed
483
            prior = self.generic_lnprior(**self.theta_prior[key])
484
485
486
487
488
489
            x = np.linspace(s.min(), s.max(), 100)
            ax2 = ax.twinx()
            ax2.get_yaxis().set_visible(False)
            ax2.plot(x, [prior(xi) for xi in x], '-r')
            ax.set_xlim(xlim)

490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
    def plot_prior_posterior(self, normal_stds=2):
        """ Plot the posterior in the context of the prior """
        fig, axes = plt.subplots(nrows=self.ndim, figsize=(8, 4*self.ndim))
        N = 1000
        from scipy.stats import gaussian_kde

        for i, (ax, key) in enumerate(zip(axes, self.theta_keys)):
            prior_dict = self.theta_prior[key]
            prior_func = self.generic_lnprior(**prior_dict)
            if prior_dict['type'] == 'unif':
                x = np.linspace(prior_dict['lower'], prior_dict['upper'], N)
                prior = prior_func(x)
                prior[0] = 0
                prior[-1] = 0
            elif prior_dict['type'] == 'norm':
                lower = prior_dict['loc'] - normal_stds * prior_dict['scale']
                upper = prior_dict['loc'] + normal_stds * prior_dict['scale']
                x = np.linspace(lower, upper, N)
                prior = prior_func(x)
509
510
511
512
513
            elif prior_dict['type'] == 'halfnorm':
                lower = prior_dict['loc']
                upper = prior_dict['loc'] + normal_stds * prior_dict['scale']
                x = np.linspace(lower, upper, N)
                prior = [prior_func(xi) for xi in x]
Gregory Ashton's avatar
Gregory Ashton committed
514
515
516
517
518
            elif prior_dict['type'] == 'neghalfnorm':
                upper = prior_dict['loc']
                lower = prior_dict['loc'] - normal_stds * prior_dict['scale']
                x = np.linspace(lower, upper, N)
                prior = [prior_func(xi) for xi in x]
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
            else:
                raise ValueError('Not implemented for prior type {}'.format(
                    prior_dict['type']))
            priorln = ax.plot(x, prior, 'r', label='prior')
            ax.set_xlabel(self.theta_symbols[i])

            s = self.samples[:, i]
            while len(s) > 10**4:
                # random downsample to avoid slow calculation of kde
                s = np.random.choice(s, size=int(len(s)/2.))
            kde = gaussian_kde(s)
            ax2 = ax.twinx()
            postln = ax2.plot(x, kde.pdf(x), 'k', label='posterior')
            ax2.set_yticklabels([])
            ax.set_yticklabels([])

        lns = priorln + postln
        labs = [l.get_label() for l in lns]
        axes[0].legend(lns, labs, loc=1, framealpha=0.8)

        fig.savefig('{}/{}_prior_posterior.png'.format(
            self.outdir, self.label))

542
    def plot_cumulative_max(self, **kwargs):
Gregory Ashton's avatar
Gregory Ashton committed
543
544
545
546
        d, maxtwoF = self.get_max_twoF()
        for key, val in self.theta_prior.iteritems():
            if key not in d:
                d[key] = val
547
548
549
550
551
552

        if hasattr(self, 'search') is False:
            self.inititate_search_object()
        if self.binary is False:
            self.search.plot_twoF_cumulative(
                self.label, self.outdir, F0=d['F0'], F1=d['F1'], F2=d['F2'],
553
                Alpha=d['Alpha'], Delta=d['Delta'],
554
                tstart=self.minStartTime, tend=self.maxStartTime,
555
                **kwargs)
556
557
558
559
560
        else:
            self.search.plot_twoF_cumulative(
                self.label, self.outdir, F0=d['F0'], F1=d['F1'], F2=d['F2'],
                Alpha=d['Alpha'], Delta=d['Delta'], asini=d['asini'],
                period=d['period'], ecc=d['ecc'], argp=d['argp'], tp=d['argp'],
561
                tstart=self.minStartTime, tend=self.maxStartTime, **kwargs)
Gregory Ashton's avatar
Gregory Ashton committed
562

Gregory Ashton's avatar
Gregory Ashton committed
563
    def generic_lnprior(self, **kwargs):
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
        """ Return a lambda function of the pdf

        Parameters
        ----------
        kwargs: dict
            A dictionary containing 'type' of pdf and shape parameters

        """

        def logunif(x, a, b):
            above = x < b
            below = x > a
            if type(above) is not np.ndarray:
                if above and below:
                    return -np.log(b-a)
                else:
                    return -np.inf
            else:
                idxs = np.array([all(tup) for tup in zip(above, below)])
                p = np.zeros(len(x)) - np.inf
                p[idxs] = -np.log(b-a)
                return p

        def halfnorm(x, loc, scale):
588
            if x < loc:
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
                return -np.inf
            else:
                return -0.5*((x-loc)**2/scale**2+np.log(0.5*np.pi*scale**2))

        def cauchy(x, x0, gamma):
            return 1.0/(np.pi*gamma*(1+((x-x0)/gamma)**2))

        def exp(x, x0, gamma):
            if x > x0:
                return np.log(gamma) - gamma*(x - x0)
            else:
                return -np.inf

        if kwargs['type'] == 'unif':
            return lambda x: logunif(x, kwargs['lower'], kwargs['upper'])
        elif kwargs['type'] == 'halfnorm':
            return lambda x: halfnorm(x, kwargs['loc'], kwargs['scale'])
606
607
        elif kwargs['type'] == 'neghalfnorm':
            return lambda x: halfnorm(-x, kwargs['loc'], kwargs['scale'])
608
609
610
611
612
613
614
        elif kwargs['type'] == 'norm':
            return lambda x: -0.5*((x - kwargs['loc'])**2/kwargs['scale']**2
                                   + np.log(2*np.pi*kwargs['scale']**2))
        else:
            logging.info("kwargs:", kwargs)
            raise ValueError("Print unrecognise distribution")

Gregory Ashton's avatar
Gregory Ashton committed
615
    def generate_rv(self, **kwargs):
616
617
618
619
620
621
622
623
        dist_type = kwargs.pop('type')
        if dist_type == "unif":
            return np.random.uniform(low=kwargs['lower'], high=kwargs['upper'])
        if dist_type == "norm":
            return np.random.normal(loc=kwargs['loc'], scale=kwargs['scale'])
        if dist_type == "halfnorm":
            return np.abs(np.random.normal(loc=kwargs['loc'],
                                           scale=kwargs['scale']))
624
625
626
        if dist_type == "neghalfnorm":
            return -1 * np.abs(np.random.normal(loc=kwargs['loc'],
                                                scale=kwargs['scale']))
627
628
629
630
631
632
        if dist_type == "lognorm":
            return np.random.lognormal(
                mean=kwargs['loc'], sigma=kwargs['scale'])
        else:
            raise ValueError("dist_type {} unknown".format(dist_type))

Gregory Ashton's avatar
Gregory Ashton committed
633
    def plot_walkers(self, sampler, symbols=None, alpha=0.4, color="k", temp=0,
634
                     lw=0.1, nprod=None, add_det_stat_burnin=False,
635
                     fig=None, axes=None, xoffset=0, plot_det_stat=True,
636
                     context='classic', subtractions=None, labelpad=0.05):
637
638
        """ Plot all the chains from a sampler """

639
640
641
        if np.ndim(axes) > 1:
            axes = axes.flatten()

642
643
644
645
646
647
648
649
650
651
652
653
654
        shape = sampler.chain.shape
        if len(shape) == 3:
            nwalkers, nsteps, ndim = shape
            chain = sampler.chain[:, :, :]
        if len(shape) == 4:
            ntemps, nwalkers, nsteps, ndim = shape
            if temp < ntemps:
                logging.info("Plotting temperature {} chains".format(temp))
            else:
                raise ValueError(("Requested temperature {} outside of"
                                  "available range").format(temp))
            chain = sampler.chain[temp, :, :, :]

655
656
        if subtractions is None:
            subtractions = [0 for i in range(ndim)]
657
658
659
        else:
            if len(subtractions) != self.ndim:
                raise ValueError('subtractions must be of length ndim')
660

661
        with plt.style.context((context)):
Gregory Ashton's avatar
Gregory Ashton committed
662
            plt.rcParams['text.usetex'] = True
Gregory Ashton's avatar
Gregory Ashton committed
663
            if fig is None and axes is None:
664
                fig = plt.figure(figsize=(4, 3.0*ndim))
Gregory Ashton's avatar
Gregory Ashton committed
665
                ax = fig.add_subplot(ndim+1, 1, 1)
Gregory Ashton's avatar
Gregory Ashton committed
666
                axes = [ax] + [fig.add_subplot(ndim+1, 1, i)
Gregory Ashton's avatar
Gregory Ashton committed
667
                               for i in range(2, ndim+1)]
668

Gregory Ashton's avatar
Gregory Ashton committed
669
            idxs = np.arange(chain.shape[1])
670
671
672
673
674
            burnin_idx = chain.shape[1] - nprod
            if hasattr(self, 'convergence_idx'):
                convergence_idx = self.convergence_idx
            else:
                convergence_idx = burnin_idx
675
676
            if ndim > 1:
                for i in range(ndim):
677
                    axes[i].ticklabel_format(useOffset=False, axis='y')
Gregory Ashton's avatar
Gregory Ashton committed
678
                    cs = chain[:, :, i].T
679
680
681
                    if burnin_idx > 0:
                        axes[i].plot(xoffset+idxs[:convergence_idx],
                                     cs[:convergence_idx]-subtractions[i],
682
                                     color="r", alpha=alpha,
Gregory Ashton's avatar
Gregory Ashton committed
683
                                     lw=lw)
684
685
                    axes[i].plot(xoffset+idxs[burnin_idx:],
                                 cs[burnin_idx:]-subtractions[i],
Gregory Ashton's avatar
Gregory Ashton committed
686
                                 color="k", alpha=alpha, lw=lw)
687
                    if symbols:
688
                        if subtractions[i] == 0:
689
                            axes[i].set_ylabel(symbols[i], labelpad=labelpad)
690
691
                        else:
                            axes[i].set_ylabel(
692
693
                                symbols[i]+'$-$'+symbols[i]+'$_0$',
                                labelpad=labelpad)
694

695
696
                    if hasattr(self, 'convergence_diagnostic'):
                        ax = axes[i].twinx()
697
698
699
700
701
                        c_x = np.array(self.convergence_diagnosticx)
                        c_y = np.array(self.convergence_diagnostic)
                        ax.plot(c_x, c_y[:, i], '-b')
                        ax.ticklabel_format(useOffset=False)
                        ax.set_ylim(1, 5)
702
            else:
Gregory Ashton's avatar
Gregory Ashton committed
703
                axes[0].ticklabel_format(useOffset=False, axis='y')
Gregory Ashton's avatar
Gregory Ashton committed
704
                cs = chain[:, :, temp].T
Gregory Ashton's avatar
Gregory Ashton committed
705
706
707
708
709
710
                if burnin_idx:
                    axes[0].plot(idxs[:burnin_idx], cs[:burnin_idx],
                                 color="r", alpha=alpha, lw=lw)
                axes[0].plot(idxs[burnin_idx:], cs[burnin_idx:], color="k",
                             alpha=alpha, lw=lw)
                if symbols:
711
                    axes[0].set_ylabel(symbols[0], labelpad=labelpad)
712

713
            if plot_det_stat:
714
715
716
                if len(axes) == ndim:
                    axes.append(fig.add_subplot(ndim+1, 1, ndim+1))

717
718
719
                lnl = sampler.lnlikelihood[temp, :, :]
                if burnin_idx and add_det_stat_burnin:
                    burn_in_vals = lnl[:, :burnin_idx].flatten()
720
721
722
723
724
725
726
                    try:
                        axes[-1].hist(burn_in_vals[~np.isnan(burn_in_vals)],
                                      bins=50, histtype='step', color='r')
                    except ValueError:
                        logging.info('Det. Stat. hist failed, most likely all '
                                     'values where the same')
                        pass
727
728
729
                else:
                    burn_in_vals = []
                prod_vals = lnl[:, burnin_idx:].flatten()
730
731
732
733
734
735
736
                try:
                    axes[-1].hist(prod_vals[~np.isnan(prod_vals)], bins=50,
                                  histtype='step', color='k')
                except ValueError:
                    logging.info('Det. Stat. hist failed, most likely all '
                                 'values where the same')
                    pass
737
738
739
740
741
742
743
744
745
746
747
748
                if self.BSGL:
                    axes[-1].set_xlabel(r'$\mathcal{B}_\mathrm{S/GL}$')
                else:
                    axes[-1].set_xlabel(r'$\widetilde{2\mathcal{F}}$')
                axes[-1].set_ylabel(r'$\textrm{Counts}$')
                combined_vals = np.append(burn_in_vals, prod_vals)
                if len(combined_vals) > 0:
                    minv = np.min(combined_vals)
                    maxv = np.max(combined_vals)
                    Range = abs(maxv-minv)
                    axes[-1].set_xlim(minv-0.1*Range, maxv+0.1*Range)

749
                xfmt = matplotlib.ticker.ScalarFormatter()
750
                xfmt.set_powerlimits((-4, 4))
751
752
                axes[-1].xaxis.set_major_formatter(xfmt)

753
            axes[-2].set_xlabel(r'$\textrm{Number of steps}$', labelpad=0.2)
754
755
        return fig, axes

Gregory Ashton's avatar
Gregory Ashton committed
756
757
758
759
760
    def apply_corrections_to_p0(self, p0):
        """ Apply any correction to the initial p0 values """
        return p0

    def generate_scattered_p0(self, p):
761
        """ Generate a set of p0s scattered about p """
Gregory Ashton's avatar
Gregory Ashton committed
762
        p0 = [[p + self.scatter_val * p * np.random.randn(self.ndim)
763
764
765
766
               for i in xrange(self.nwalkers)]
              for j in xrange(self.ntemps)]
        return p0

Gregory Ashton's avatar
Gregory Ashton committed
767
    def generate_initial_p0(self):
768
769
770
        """ Generate a set of init vals for the walkers """

        if type(self.theta_initial) == dict:
771
            logging.info('Generate initial values from initial dictionary')
772
            if hasattr(self, 'nglitch') and self.nglitch > 1:
773
                raise ValueError('Initial dict not implemented for nglitch>1')
Gregory Ashton's avatar
Gregory Ashton committed
774
            p0 = [[[self.generate_rv(**self.theta_initial[key])
775
776
777
                    for key in self.theta_keys]
                   for i in range(self.nwalkers)]
                  for j in range(self.ntemps)]
778
779
780
781
782
783
        elif type(self.theta_initial) == list:
            logging.info('Generate initial values from list of theta_initial')
            p0 = [[[self.generate_rv(**val)
                    for val in self.theta_initial]
                   for i in range(self.nwalkers)]
                  for j in range(self.ntemps)]
784
        elif self.theta_initial is None:
785
            logging.info('Generate initial values from prior dictionary')
Gregory Ashton's avatar
Gregory Ashton committed
786
            p0 = [[[self.generate_rv(**self.theta_prior[key])
787
788
789
790
                    for key in self.theta_keys]
                   for i in range(self.nwalkers)]
                  for j in range(self.ntemps)]
        elif len(self.theta_initial) == self.ndim:
Gregory Ashton's avatar
Gregory Ashton committed
791
            p0 = self.generate_scattered_p0(self.theta_initial)
792
793
794
795
796
        else:
            raise ValueError('theta_initial not understood')

        return p0

797
    def get_new_p0(self, sampler):
798
799
800
801
802
803
        """ Returns new initial positions for walkers are burn0 stage

        This returns new positions for all walkers by scattering points about
        the maximum posterior with scale `scatter_val`.

        """
Gregory Ashton's avatar
Gregory Ashton committed
804
805
806
807
        temp_idx = 0
        pF = sampler.chain[temp_idx, :, :, :]
        lnl = sampler.lnlikelihood[temp_idx, :, :]
        lnp = sampler.lnprobability[temp_idx, :, :]
808
809

        # General warnings about the state of lnp
Gregory Ashton's avatar
Gregory Ashton committed
810
        if np.any(np.isnan(lnp)):
811
812
            logging.warning(
                "Of {} lnprobs {} are nan".format(
Gregory Ashton's avatar
Gregory Ashton committed
813
814
                    np.shape(lnp), np.sum(np.isnan(lnp))))
        if np.any(np.isposinf(lnp)):
815
816
            logging.warning(
                "Of {} lnprobs {} are +np.inf".format(
Gregory Ashton's avatar
Gregory Ashton committed
817
818
                    np.shape(lnp), np.sum(np.isposinf(lnp))))
        if np.any(np.isneginf(lnp)):
819
820
            logging.warning(
                "Of {} lnprobs {} are -np.inf".format(
Gregory Ashton's avatar
Gregory Ashton committed
821
                    np.shape(lnp), np.sum(np.isneginf(lnp))))
822

823
824
        lnp_finite = copy.copy(lnp)
        lnp_finite[np.isinf(lnp)] = np.nan
Gregory Ashton's avatar
Gregory Ashton committed
825
826
        idx = np.unravel_index(np.nanargmax(lnp_finite), lnp_finite.shape)
        p = pF[idx]
827
        p0 = self.generate_scattered_p0(p)
828

829
830
831
832
833
834
835
836
        self.search.BSGL = False
        twoF = self.logl(p, self.search)
        self.search.BSGL = self.BSGL

        logging.info(('Gen. new p0 from pos {} which had det. stat.={:2.1f},'
                      ' twoF={:2.1f} and lnp={:2.1f}')
                     .format(idx[1], lnl[idx], twoF, lnp_finite[idx]))

837
838
839
840
841
        return p0

    def get_save_data_dictionary(self):
        d = dict(nsteps=self.nsteps, nwalkers=self.nwalkers,
                 ntemps=self.ntemps, theta_keys=self.theta_keys,
Gregory Ashton's avatar
Gregory Ashton committed
842
                 theta_prior=self.theta_prior, scatter_val=self.scatter_val,
843
                 log10temperature_min=self.log10temperature_min,
844
                 BSGL=self.BSGL)
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
        return d

    def save_data(self, sampler, samples, lnprobs, lnlikes):
        d = self.get_save_data_dictionary()
        d['sampler'] = sampler
        d['samples'] = samples
        d['lnprobs'] = lnprobs
        d['lnlikes'] = lnlikes

        if os.path.isfile(self.pickle_path):
            logging.info('Saving backup of {} as {}.old'.format(
                self.pickle_path, self.pickle_path))
            os.rename(self.pickle_path, self.pickle_path+".old")
        with open(self.pickle_path, "wb") as File:
            pickle.dump(d, File)

    def get_saved_data(self):
        with open(self.pickle_path, "r") as File:
            d = pickle.load(File)
        return d

    def check_old_data_is_okay_to_use(self):
867
868
869
870
        if args.use_old_data:
            logging.info("Forcing use of old data")
            return True

871
872
873
874
        if os.path.isfile(self.pickle_path) is False:
            logging.info('No pickled data found')
            return False

Gregory Ashton's avatar
Gregory Ashton committed
875
876
877
878
879
880
        if self.sftfilepath is not None:
            oldest_sft = min([os.path.getmtime(f) for f in
                              self.get_list_of_matching_sfts()])
            if os.path.getmtime(self.pickle_path) < oldest_sft:
                logging.info('Pickled data outdates sft files')
                return False
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895

        old_d = self.get_saved_data().copy()
        new_d = self.get_save_data_dictionary().copy()

        old_d.pop('samples')
        old_d.pop('sampler')
        old_d.pop('lnprobs')
        old_d.pop('lnlikes')

        mod_keys = []
        for key in new_d.keys():
            if key in old_d:
                if new_d[key] != old_d[key]:
                    mod_keys.append((key, old_d[key], new_d[key]))
            else:
896
                raise ValueError('Keys {} not in old dictionary'.format(key))
897
898
899
900
901
902
903
904
905

        if len(mod_keys) == 0:
            return True
        else:
            logging.warning("Saved data differs from requested")
            logging.info("Differences found in following keys:")
            for key in mod_keys:
                if len(key) == 3:
                    if np.isscalar(key[1]) or key[0] == 'nsteps':
906
                        logging.info("    {} : {} -> {}".format(*key))
907
                    else:
908
                        logging.info("    " + key[0])
909
910
911
912
913
                else:
                    logging.info(key)
            return False

    def get_max_twoF(self, threshold=0.05):
914
        """ Returns the max likelihood sample and the corresponding 2F value
915
916
917
918
919
920
921
922
923
924
925
926
927
928

        Note: the sample is returned as a dictionary along with an estimate of
        the standard deviation calculated from the std of all samples with a
        twoF within `threshold` (relative) to the max twoF

        """
        if any(np.isposinf(self.lnlikes)):
            logging.info('twoF values contain positive infinite values')
        if any(np.isneginf(self.lnlikes)):
            logging.info('twoF values contain negative infinite values')
        if any(np.isnan(self.lnlikes)):
            logging.info('twoF values contain nan')
        idxs = np.isfinite(self.lnlikes)
        jmax = np.nanargmax(self.lnlikes[idxs])
929
        maxlogl = self.lnlikes[jmax]
930
        d = OrderedDict()
931

932
933
934
935
936
937
938
939
940
941
        if self.BSGL:
            if hasattr(self, 'search') is False:
                self.inititate_search_object()
            p = self.samples[jmax]
            self.search.BSGL = False
            maxtwoF = self.logl(p, self.search)
            self.search.BSGL = self.BSGL
        else:
            maxtwoF = maxlogl

Gregory Ashton's avatar
Gregory Ashton committed
942
        repeats = []
943
        for i, k in enumerate(self.theta_keys):
Gregory Ashton's avatar
Gregory Ashton committed
944
945
946
947
948
949
950
951
952
953
            if k in d and k not in repeats:
                d[k+'_0'] = d[k]  # relabel the old key
                d.pop(k)
                repeats.append(k)
            if k in repeats:
                k = k + '_0'
                count = 1
                while k in d:
                    k = k.replace('_{}'.format(count-1), '_{}'.format(count))
                    count += 1
954
955
956
957
958
            d[k] = self.samples[jmax][i]
        return d, maxtwoF

    def get_median_stds(self):
        """ Returns a dict of the median and std of all production samples """
959
        d = OrderedDict()
Gregory Ashton's avatar
Gregory Ashton committed
960
        repeats = []
961
        for s, k in zip(self.samples.T, self.theta_keys):
Gregory Ashton's avatar
Gregory Ashton committed
962
963
964
965
966
967
968
969
970
971
972
973
974
            if k in d and k not in repeats:
                d[k+'_0'] = d[k]  # relabel the old key
                d[k+'_0_std'] = d[k+'_std']
                d.pop(k)
                d.pop(k+'_std')
                repeats.append(k)
            if k in repeats:
                k = k + '_0'
                count = 1
                while k in d:
                    k = k.replace('_{}'.format(count-1), '_{}'.format(count))
                    count += 1

975
976
977
978
            d[k] = np.median(s)
            d[k+'_std'] = np.std(s)
        return d

979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
    def check_if_samples_are_railing(self, threshold=0.01):
        return_flag = False
        for s, k in zip(self.samples.T, self.theta_keys):
            prior = self.theta_prior[k]
            if prior['type'] == 'unif':
                prior_range = prior['upper'] - prior['lower']
                edges = []
                fracs = []
                for l in ['lower', 'upper']:
                    bools = np.abs(s - prior[l])/prior_range < threshold
                    if np.any(bools):
                        edges.append(l)
                        fracs.append(str(100*float(np.sum(bools))/len(bools)))
                if len(edges) > 0:
                    logging.warning(
                        '{}% of the {} posterior is railing on the {} edges'
                        .format('% & '.join(fracs), k, ' & '.join(edges)))
                    return_flag = True
        return return_flag

999
1000
1001
1002
    def write_par(self, method='med'):
        """ Writes a .par of the best-fit params with an estimated std """
        logging.info('Writing {}/{}.par using the {} method'.format(
            self.outdir, self.label, method))
1003
1004
1005
1006

        median_std_d = self.get_median_stds()
        max_twoF_d, max_twoF = self.get_max_twoF()

Gregory Ashton's avatar
Gregory Ashton committed
1007
        logging.info('Writing par file with max twoF = {}'.format(max_twoF))
1008
1009
1010
        filename = '{}/{}.par'.format(self.outdir, self.label)
        with open(filename, 'w+') as f:
            f.write('MaxtwoF = {}\n'.format(max_twoF))
Gregory Ashton's avatar
Gregory Ashton committed
1011
            f.write('tref = {}\n'.format(self.tref))
1012
1013
            if hasattr(self, 'theta0_index'):
                f.write('theta0_index = {}\n'.format(self.theta0_idx))
1014
            if method == 'med':
1015
1016
                for key, val in median_std_d.iteritems():
                    f.write('{} = {:1.16e}\n'.format(key, val))
1017
            if method == 'twoFmax':
1018
1019
1020
1021
                for key, val in max_twoF_d.iteritems():
                    f.write('{} = {:1.16e}\n'.format(key, val))

    def print_summary(self):
Gregory Ashton's avatar
Gregory Ashton committed
1022
        max_twoFd, max_twoF = self.get_max_twoF()
1023
        median_std_d = self.get_median_stds()
Gregory Ashton's avatar
Gregory Ashton committed
1024
        logging.info('Summary:')
1025
        if hasattr(self, 'theta0_idx'):
Gregory Ashton's avatar
Gregory Ashton committed
1026
1027
            logging.info('theta0 index: {}'.format(self.theta0_idx))
        logging.info('Max twoF: {} with parameters:'.format(max_twoF))
Gregory Ashton's avatar
Gregory Ashton committed
1028
1029
        for k in np.sort(max_twoFd.keys()):
            print('  {:10s} = {:1.9e}'.format(k, max_twoFd[k]))
Gregory Ashton's avatar
Gregory Ashton committed
1030
        logging.info('Median +/- std for production values')
1031
        for k in np.sort(median_std_d.keys()):
1032
            if 'std' not in k:
Gregory Ashton's avatar
Gregory Ashton committed
1033
                logging.info('  {:10s} = {:1.9e} +/- {:1.9e}'.format(
1034
                    k, median_std_d[k], median_std_d[k+'_std']))
Gregory Ashton's avatar
Gregory Ashton committed
1035
        logging.info('\n')
1036

1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
    def CF_twoFmax(self, theta, twoFmax, ntrials):
        Fmax = twoFmax/2.0
        return (np.exp(1j*theta*twoFmax)*ntrials/2.0
                * Fmax*np.exp(-Fmax)*(1-(1+Fmax)*np.exp(-Fmax))**(ntrials-1))

    def pdf_twoFhat(self, twoFhat, nglitch, ntrials, twoFmax=100, dtwoF=0.1):
        if np.ndim(ntrials) == 0:
            ntrials = np.zeros(nglitch+1) + ntrials
        twoFmax_int = np.arange(0, twoFmax, dtwoF)
        theta_int = np.arange(-1/dtwoF, 1./dtwoF, 1./twoFmax)
        CF_twoFmax_theta = np.array(
            [[np.trapz(self.CF_twoFmax(t, twoFmax_int, ntrial), twoFmax_int)
              for t in theta_int]
             for ntrial in ntrials])
        CF_twoFhat_theta = np.prod(CF_twoFmax_theta, axis=0)
        pdf = (1/(2*np.pi)) * np.array(
            [np.trapz(np.exp(-1j*theta_int*twoFhat_val)
             * CF_twoFhat_theta, theta_int) for twoFhat_val in twoFhat])
        return pdf.real

    def p_val_twoFhat(self, twoFhat, ntrials, twoFhatmax=500, Npoints=1000):
1058
        """ Caluculate the p-value for the given twoFhat in Gaussian noise
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
1072
1073
1074
1075
1076
1077

        Parameters
        ----------
        twoFhat: float
            The observed twoFhat value
        ntrials: int, array of len Nglitch+1
            The number of trials for each glitch+1
        """
        twoFhats = np.linspace(twoFhat, twoFhatmax, Npoints)
        pdf = self.pdf_twoFhat(twoFhats, self.nglitch, ntrials)
        return np.trapz(pdf, twoFhats)

    def get_p_value(self, delta_F0, time_trials=0):
        """ Get's the p-value for the maximum twoFhat value """
        d, max_twoF = self.get_max_twoF()
        if self.nglitch == 1:
            tglitches = [d['tglitch']]
        else:
            tglitches = [d['tglitch_{}'.format(i)] for i in range(self.nglitch)]
1078
        tboundaries = [self.minStartTime] + tglitches + [self.maxStartTime]
1079
        deltaTs = np.diff(tboundaries)
1080
1081
        ntrials = [time_trials + delta_F0 * dT for dT in deltaTs]
        p_val = self.p_val_twoFhat(max_twoF, ntrials)
1082
        print('p-value = {}'.format(p_val))
1083
1084
        return p_val

1085
    def get_evidence(self):
1086
1087
1088
1089
1090
1091
        fburnin = float(self.nsteps[-2])/np.sum(self.nsteps[-2:])
        lnev, lnev_err = self.sampler.thermodynamic_integration_log_evidence(
            fburnin=fburnin)

        log10evidence = lnev/np.log(10)
        log10evidence_err = lnev_err/np.log(10)
1092
1093
1094
1095
        return log10evidence, log10evidence_err

    def compute_evidence_long(self):
        """ Computes the evidence/marginal likelihood for the model """
1096
1097
1098
1099
1100
1101
1102
1103
1104
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116
1117
1118
1119
1120
1121
1122
1123
1124
1125
1126
1127
1128
1129
1130
1131
1132
1133
1134
1135
        betas = self.betas
        alllnlikes = self.sampler.lnlikelihood[:, :, self.nsteps[-2]:]
        mean_lnlikes = np.mean(np.mean(alllnlikes, axis=1), axis=1)

        mean_lnlikes = mean_lnlikes[::-1]
        betas = betas[::-1]

        fig, (ax1, ax2) = plt.subplots(nrows=2, figsize=(6, 8))

        if any(np.isinf(mean_lnlikes)):
            print("WARNING mean_lnlikes contains inf: recalculating without"
                  " the {} infs".format(len(betas[np.isinf(mean_lnlikes)])))
            idxs = np.isinf(mean_lnlikes)
            mean_lnlikes = mean_lnlikes[~idxs]
            betas = betas[~idxs]
            log10evidence = np.trapz(mean_lnlikes, betas)/np.log(10)
            z1 = np.trapz(mean_lnlikes, betas)
            z2 = np.trapz(mean_lnlikes[::-1][::2][::-1],
                          betas[::-1][::2][::-1])
            log10evidence_err = np.abs(z1 - z2) / np.log(10)

        ax1.semilogx(betas, mean_lnlikes, "-o")
        ax1.set_xlabel(r"$\beta$")
        ax1.set_ylabel(r"$\langle \log(\mathcal{L}) \rangle$")
        print("log10 evidence for {} = {} +/- {}".format(
              self.label, log10evidence, log10evidence_err))
        min_betas = []
        evidence = []
        for i in range(len(betas)/2):
            min_betas.append(betas[i])
            lnZ = np.trapz(mean_lnlikes[i:], betas[i:])
            evidence.append(lnZ/np.log(10))

        ax2.semilogx(min_betas, evidence, "-o")
        ax2.set_ylabel(r"$\int_{\beta_{\textrm{Min}}}^{\beta=1}" +
                       r"\langle \log(\mathcal{L})\rangle d\beta$", size=16)
        ax2.set_xlabel(r"$\beta_{\textrm{min}}$")
        plt.tight_layout()
        fig.savefig("{}/{}_beta_lnl.png".format(self.outdir, self.label))

1136

Gregory Ashton's avatar
Gregory Ashton committed
1137
1138
class MCMCGlitchSearch(MCMCSearch):
    """ MCMC search using the SemiCoherentGlitchSearch """
Gregory Ashton's avatar
Gregory Ashton committed
1139
    @helper_functions.initializer
1140
    def __init__(self, label, outdir, sftfilepath, theta_prior, tref,
Gregory Ashton's avatar
Gregory Ashton committed
1141
                 minStartTime, maxStartTime, nglitch=1, nsteps=[100, 100],
1142
1143
                 nwalkers=100, ntemps=1, log10temperature_min=-5,
                 theta_initial=None, scatter_val=1e-10, dtglitchmin=1*86400,
1144
                 theta0_idx=0, detectors=None, BSGL=False, minCoverFreq=None,
1145
                 maxCoverFreq=None, earth_ephem=None, sun_ephem=None):
Gregory Ashton's avatar
Gregory Ashton committed
1146
1147
        """
        Parameters
Gregory Ashton's avatar
Gregory Ashton committed
1148
        ----------
Gregory Ashton's avatar
Gregory Ashton committed
1149
1150
        label, outdir: str
            A label and directory to read/write data from/to
Gregory Ashton's avatar
Gregory Ashton committed
1151
        sftfilepath: str
1152
            File patern to match SFTs
Gregory Ashton's avatar
Gregory Ashton committed
1153
1154
1155
1156
1157
1158
1159
1160
        theta_prior: dict
            Dictionary of priors and fixed values for the search parameters.
            For each parameters (key of the dict), if it is to be held fixed
            the value should be the constant float, if it is be searched, the
            value should be a dictionary of the prior.
        theta_initial: dict, array, (None)
            Either a dictionary of distribution about which to distribute the
            initial walkers about, an array (from which the walkers will be
Gregory Ashton's avatar
Gregory Ashton committed
1161
            scattered by scatter_val), or None in which case the prior is used.
1162
1163
1164
1165
        scatter_val, float or ndim array
            Size of scatter to use about the initialisation step, if given as
            an array it must be of length ndim and the order is given by
            theta_keys
Gregory Ashton's avatar
Gregory Ashton committed
1166
1167
        nglitch: int
            The number of glitches to allow
1168
        tref, minStartTime, maxStartTime: int
Gregory Ashton's avatar
Gregory Ashton committed
1169
1170
1171
1172
1173
1174
1175
1176
1177
            GPS seconds of the reference time, start time and end time
        nsteps: list (m,)
            List specifying the number of steps to take, the last two entries
            give the nburn and nprod of the 'production' run, all entries
            before are for iterative initialisation steps (usually just one)
            e.g. [1000, 1000, 500].
        dtglitchmin: int
            The minimum duration (in seconds) of a segment between two glitches
            or a glitch and the start/end of the data
1178
1179
1180
1181
1182
1183
        nwalkers, ntemps: int,
            The number of walkers and temperates to use in the parallel
            tempered PTSampler.
        log10temperature_min float < 0
            The  log_10(tmin) value, the set of betas passed to PTSampler are
            generated from np.logspace(0, log10temperature_min, ntemps).
1184
1185
1186
1187
        theta0_idx, int
            Index (zero-based) of which segment the theta refers to - uyseful
            if providing a tight prior on theta to allow the signal to jump
            too theta (and not just from)
1188
        detectors: str
1189
1190
            Two character reference to the data to use, specify None for no
            contraint.
Gregory Ashton's avatar
Gregory Ashton committed
1191
1192
1193
1194
1195
1196
1197
1198
1199
1200
        minCoverFreq, maxCoverFreq: float
            Minimum and maximum instantaneous frequency which will be covered
            over the SFT time span as passed to CreateFstatInput
        earth_ephem, sun_ephem: str
            Paths of the two files containing positions of Earth and Sun,
            respectively at evenly spaced times, as passed to CreateFstatInput
            If None defaults defined in BaseSearchClass will be used

        """

Gregory Ashton's avatar
Gregory Ashton committed
1201
1202
        if os.path.isdir(outdir) is False:
            os.mkdir(outdir)
1203
        self.add_log_file()
Gregory Ashton's avatar
Gregory Ashton committed
1204
1205
        logging.info(('Set-up MCMC glitch search with {} glitches for model {}'
                      ' on data {}').format(self.nglitch, self.label,
1206
                                            self.sftfilepath))
Gregory Ashton's avatar
Gregory Ashton committed
1207
1208
1209
        self.pickle_path = '{}/{}_saved_data.p'.format(self.outdir, self.label)
        self.unpack_input_theta()
        self.ndim = len(self.theta_keys)
1210
1211
1212
1213
        if self.log10temperature_min:
            self.betas = np.logspace(0, self.log10temperature_min, self.ntemps)
        else:
            self.betas = None
Gregory Ashton's avatar
Gregory Ashton committed
1214
1215
1216
1217
1218
1219
1220
1221
1222
        if earth_ephem is None:
            self.earth_ephem = self.earth_ephem_default
        if sun_ephem is None:
            self.sun_ephem = self.sun_ephem_default

        if args.clean and os.path.isfile(self.pickle_path):
            os.rename(self.pickle_path, self.pickle_path+".old")

        self.old_data_is_okay_to_use = self.check_old_data_is_okay_to_use()
1223
        self.log_input()
Gregory Ashton's avatar
Gregory Ashton committed
1224
1225
1226
1227

    def inititate_search_object(self):
        logging.info('Setting up search object')
        self.search = SemiCoherentGlitchSearch(
1228
            label=self.label, outdir=self.outdir, sftfilepath=self.sftfilepath,
1229
1230
            tref=self.tref, minStartTime=self.minStartTime,
            maxStartTime=self.maxStartTime, minCoverFreq=self.minCoverFreq,
Gregory Ashton's avatar
Gregory Ashton committed
1231
            maxCoverFreq=self.maxCoverFreq, earth_ephem=self.earth_ephem,
1232
            sun_ephem=self.sun_ephem, detectors=self.detectors, BSGL=self.BSGL,
1233
            nglitch=self.nglitch, theta0_idx=self.theta0_idx)
Gregory Ashton's avatar
Gregory Ashton committed