pyfstat.py 63.8 KB
Newer Older
1
2
3
4
5
6
7
8
9
10
""" Classes for various types of searches using ComputeFstatistic """
import os
import sys
import itertools
import logging
import argparse
import copy
import glob
import inspect
from functools import wraps
11
import subprocess
12
from collections import OrderedDict
13
14
15
16
17
18
19
20
21
22

import numpy as np
import matplotlib
matplotlib.use('Agg')
import matplotlib.pyplot as plt
import emcee
import corner
import dill as pickle
import lalpulsar

23
plt.rcParams['text.usetex'] = True
24
plt.rcParams['axes.formatter.useoffset'] = False
25

26
27
28
29
30
31
32
config_file = os.path.expanduser('~')+'/.pyfstat.conf'
if os.path.isfile(config_file):
    d = {}
    with open(config_file, 'r') as f:
        for line in f:
            k, v = line.split('=')
            k = k.replace(' ', '')
33
            v = v.replace(' ', '').replace("'", "").replace('"', '').replace('\n', '')
34
35
36
37
38
39
            d[k] = v
    earth_ephem = d['earth_ephem']
    sun_ephem = d['sun_ephem']
else:
    logging.warning('No ~/.pyfstat.conf file found please provide the paths '
                    'when initialising searches')
40
41
42
    earth_ephem = None
    sun_ephem = None

43
44
45
46
47
parser = argparse.ArgumentParser()
parser.add_argument("-q", "--quite", help="Decrease output verbosity",
                    action="store_true")
parser.add_argument("-c", "--clean", help="Don't use cached data",
                    action="store_true")
48
parser.add_argument("-u", "--use-old-data", action="store_true")
49
50
51
52
53
54
55
56
57
58
59
60
61
parser.add_argument('unittest_args', nargs='*')
args, unknown = parser.parse_known_args()
sys.argv[1:] = args.unittest_args

if args.quite:
    log_level = logging.WARNING
else:
    log_level = logging.DEBUG

logging.basicConfig(level=log_level,
                    format='%(asctime)s %(levelname)-8s: %(message)s',
                    datefmt='%H:%M')

62
63

def initializer(func):
64
    """ Automatically assigns the parameters to self """
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
    names, varargs, keywords, defaults = inspect.getargspec(func)

    @wraps(func)
    def wrapper(self, *args, **kargs):
        for name, arg in list(zip(names[1:], args)) + list(kargs.items()):
            setattr(self, name, arg)

        for name, default in zip(reversed(names), reversed(defaults)):
            if not hasattr(self, name):
                setattr(self, name, default)

        func(self, *args, **kargs)

    return wrapper


def read_par(label, outdir):
82
    """ Read in a .par file, returns a dictionary of the values """
83
84
85
86
87
88
89
90
91
92
    filename = '{}/{}.par'.format(outdir, label)
    d = {}
    with open(filename, 'r') as f:
        for line in f:
            key, val = line.rstrip('\n').split(' = ')
            d[key] = np.float64(val)
    return d


class BaseSearchClass(object):
93
    """ The base search class, provides ephemeris and general utilities """
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122

    earth_ephem_default = earth_ephem
    sun_ephem_default = sun_ephem

    def shift_matrix(self, n, dT):
        """ Generate the shift matrix """
        m = np.zeros((n, n))
        factorial = np.math.factorial
        for i in range(n):
            for j in range(n):
                if i == j:
                    m[i, j] = 1.0
                elif i > j:
                    m[i, j] = 0.0
                else:
                    if i == 0:
                        m[i, j] = 2*np.pi*float(dT)**(j-i) / factorial(j-i)
                    else:
                        m[i, j] = float(dT)**(j-i) / factorial(j-i)

        return m

    def shift_coefficients(self, theta, dT):
        """ Shift a set of coefficients by dT

        Parameters
        ----------
        theta: array-like, shape (n,)
            vector of the expansion coefficients to transform starting from the
123
            lowest degree e.g [phi, F0, F1,...].
124
        dT: float
125
            difference between the two reference times as tref_new - tref_old.
126
127
128
129

        Returns
        -------
        theta_new: array-like shape (n,)
130
            vector of the coefficients as evaluate as the new reference time.
131
132
133
134
135
        """
        n = len(theta)
        m = self.shift_matrix(n, dT)
        return np.dot(m, theta)

136
    def calculate_thetas(self, theta, delta_thetas, tbounds, theta0_idx=0):
137
138
139
        """ Calculates the set of coefficients for the post-glitch signal """
        thetas = [theta]
        for i, dt in enumerate(delta_thetas):
140
141
142
143
144
145
146
147
148
149
150
151
152
            if i < theta0_idx:
                pre_theta_at_ith_glitch = self.shift_coefficients(
                    thetas[0], tbounds[i+1] - self.tref)
                post_theta_at_ith_glitch = pre_theta_at_ith_glitch - dt
                thetas.insert(0, self.shift_coefficients(
                    post_theta_at_ith_glitch, self.tref - tbounds[i+1]))

            elif i >= theta0_idx:
                pre_theta_at_ith_glitch = self.shift_coefficients(
                    thetas[i], tbounds[i+1] - self.tref)
                post_theta_at_ith_glitch = pre_theta_at_ith_glitch + dt
                thetas.append(self.shift_coefficients(
                    post_theta_at_ith_glitch, self.tref - tbounds[i+1]))
153
154
155
        return thetas


Gregory Ashton's avatar
Gregory Ashton committed
156
157
158
159
160
161
162
163
164
class ComputeFstat(object):
    """ Base class providing interface to lalpulsar.ComputeFstat """

    earth_ephem_default = earth_ephem
    sun_ephem_default = sun_ephem

    @initializer
    def __init__(self, tref, sftlabel=None, sftdir=None,
                 minCoverFreq=None, maxCoverFreq=None,
165
166
                 detector=None, earth_ephem=None, sun_ephem=None,
                 binary=False, transient=True):
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
        """
        Parameters
        ----------
        tref: int
            GPS seconds of the reference time.
        sftlabel, sftdir: str
            A label and directory in which to find the relevant sft file
        minCoverFreq, maxCoverFreq: float
            The min and max cover frequency passed to CreateFstatInput, if
            either is None the range of frequencies in the SFT less 1Hz is
            used.
        detector: str
            Two character reference to the data to use, specify None for no
            contraint.
        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.
        binary: bool
            If true, search of binary parameters.
        transient: bool
            If true, allow for the Fstat to be computed over a transient range.

        """
Gregory Ashton's avatar
Gregory Ashton committed
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208

        if earth_ephem is None:
            self.earth_ephem = self.earth_ephem_default
        if sun_ephem is None:
            self.sun_ephem = self.sun_ephem_default

        self.init_computefstatistic_single_point()

    def init_computefstatistic_single_point(self):
        """ Initilisation step of run_computefstatistic for a single point """

        logging.info('Initialising SFTCatalog')
        constraints = lalpulsar.SFTConstraints()
        if self.detector:
            constraints.detector = self.detector
        self.sft_filepath = self.sftdir+'/*_'+self.sftlabel+"*sft"
        SFTCatalog = lalpulsar.SFTdataFind(self.sft_filepath, constraints)
        names = list(set([d.header.name for d in SFTCatalog.data]))
209
210
211
        logging.info(
            'Loaded data from detectors {} matching pattern {}'.format(
                names, self.sft_filepath))
Gregory Ashton's avatar
Gregory Ashton committed
212
213
214
215
216
217

        logging.info('Initialising ephems')
        ephems = lalpulsar.InitBarycenter(self.earth_ephem, self.sun_ephem)

        logging.info('Initialising FstatInput')
        dFreq = 0
218
219
220
221
222
        if self.transient:
            self.whatToCompute = lalpulsar.FSTATQ_ATOMS_PER_DET
        else:
            self.whatToCompute = lalpulsar.FSTATQ_2F

Gregory Ashton's avatar
Gregory Ashton committed
223
224
225
226
227
228
229
230
        FstatOptionalArgs = lalpulsar.FstatOptionalArgsDefaults

        if self.minCoverFreq is None or self.maxCoverFreq is None:
            fA = SFTCatalog.data[0].header.f0
            numBins = SFTCatalog.data[0].numBins
            fB = fA + (numBins-1)*SFTCatalog.data[0].header.deltaF
            self.minCoverFreq = fA + 0.5
            self.maxCoverFreq = fB - 0.5
231
232
233
            logging.info('Min/max cover freqs not provided, using '
                         '{} and {}, est. from SFTs'.format(
                             self.minCoverFreq, self.maxCoverFreq))
Gregory Ashton's avatar
Gregory Ashton committed
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253

        self.FstatInput = lalpulsar.CreateFstatInput(SFTCatalog,
                                                     self.minCoverFreq,
                                                     self.maxCoverFreq,
                                                     dFreq,
                                                     ephems,
                                                     FstatOptionalArgs
                                                     )

        logging.info('Initialising PulsarDoplerParams')
        PulsarDopplerParams = lalpulsar.PulsarDopplerParams()
        PulsarDopplerParams.refTime = self.tref
        PulsarDopplerParams.Alpha = 1
        PulsarDopplerParams.Delta = 1
        PulsarDopplerParams.fkdot = np.array([0, 0, 0, 0, 0, 0, 0])
        self.PulsarDopplerParams = PulsarDopplerParams

        logging.info('Initialising FstatResults')
        self.FstatResults = lalpulsar.FstatResults()

254
255
256
257
258
259
260
        if self.transient:
            self.windowRange = lalpulsar.transientWindowRange_t()
            self.windowRange.type = lalpulsar.TRANSIENT_RECTANGULAR
            self.windowRange.t0Band = 0
            self.windowRange.dt0 = 1
            self.windowRange.tauBand = 0
            self.windowRange.dtau = 1
261

Gregory Ashton's avatar
Gregory Ashton committed
262
    def run_computefstatistic_single_point(self, tstart, tend, F0, F1,
263
264
265
                                           F2, Alpha, Delta, asini=None,
                                           period=None, ecc=None, tp=None,
                                           argp=None):
266
        """ Returns the twoF fully-coherently at a single point """
Gregory Ashton's avatar
Gregory Ashton committed
267
268
269
270

        self.PulsarDopplerParams.fkdot = np.array([F0, F1, F2, 0, 0, 0, 0])
        self.PulsarDopplerParams.Alpha = Alpha
        self.PulsarDopplerParams.Delta = Delta
271
272
273
274
275
276
        if self.binary:
            self.PulsarDopplerParams.asini = asini
            self.PulsarDopplerParams.period = period
            self.PulsarDopplerParams.ecc = ecc
            self.PulsarDopplerParams.tp = tp
            self.PulsarDopplerParams.argp = argp
Gregory Ashton's avatar
Gregory Ashton committed
277
278
279
280

        lalpulsar.ComputeFstat(self.FstatResults,
                               self.FstatInput,
                               self.PulsarDopplerParams,
281
                               1,
Gregory Ashton's avatar
Gregory Ashton committed
282
283
284
                               self.whatToCompute
                               )

285
286
287
        if self.transient is False:
            return self.FstatResults.twoF[0]

288
289
        self.windowRange.t0 = int(tstart)  # TYPE UINT4
        self.windowRange.tau = int(tend - tstart)  # TYPE UINT4
Gregory Ashton's avatar
Gregory Ashton committed
290
        FS = lalpulsar.ComputeTransientFstatMap(
291
            self.FstatResults.multiFatoms[0], self.windowRange, False)
Gregory Ashton's avatar
Gregory Ashton committed
292
293
294
295
        return 2*FS.F_mn.data[0][0]


class SemiCoherentGlitchSearch(BaseSearchClass, ComputeFstat):
296
297
298
299
300
301
302
303
304
    """ A semi-coherent glitch search

    This implements a basic `semi-coherent glitch F-stat in which the data
    is divided into two segments either side of the proposed glitch and the
    fully-coherent F-stat in each segment is averaged to give the semi-coherent
    F-stat
    """

    @initializer
Gregory Ashton's avatar
Gregory Ashton committed
305
    def __init__(self, label, outdir, tref, tstart, tend, nglitch=0,
306
                 sftlabel=None, sftdir=None, theta0_idx=0, minCoverFreq=None,
307
308
                 maxCoverFreq=None, detector=None, earth_ephem=None,
                 sun_ephem=None):
309
310
311
312
        """
        Parameters
        ----------
        label, outdir: str
313
314
315
316
317
318
            A label and directory to read/write data from/to.
        tref, tstart, tend: int
            GPS seconds of the reference time, and start and end of the data.
        nglitch: int
            The (fixed) number of glitches; this can zero, but occasionally
            this causes issue (in which case just use ComputeFstat).
319
320
        sftlabel, sftdir: str
            A label and directory in which to find the relevant sft file. If
321
            None use label and outdir.
322
323
324
325
        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)
326
        minCoverFreq, maxCoverFreq: float
327
328
329
            The min and max cover frequency passed to CreateFstatInput, if
            either is None the range of frequencies in the SFT less 1Hz is
            used.
330
331
        detector: str
            Two character reference to the data to use, specify None for no
332
            contraint.
333
334
        earth_ephem, sun_ephem: str
            Paths of the two files containing positions of Earth and Sun,
335
336
            respectively at evenly spaced times, as passed to CreateFstatInput.
            If None defaults defined in BaseSearchClass will be used.
337
338
339
340
341
342
343
344
345
346
347
        """

        if self.sftlabel is None:
            self.sftlabel = self.label
        if self.sftdir is None:
            self.sftdir = self.outdir
        self.fs_file_name = "{}/{}_FS.dat".format(self.outdir, self.label)
        if self.earth_ephem is None:
            self.earth_ephem = self.earth_ephem_default
        if self.sun_ephem is None:
            self.sun_ephem = self.sun_ephem_default
348
349
        self.transient = True
        self.binary = False
350
351
352
        self.init_computefstatistic_single_point()

    def compute_nglitch_fstat(self, F0, F1, F2, Alpha, Delta, *args):
353
        """ Returns the semi-coherent glitch summed twoF """
354
355
356

        args = list(args)
        tboundaries = [self.tstart] + args[-self.nglitch:] + [self.tend]
357
358
359
360
361
362
363
364
        delta_F0s = args[-3*self.nglitch:-2*self.nglitch]
        delta_F1s = args[-2*self.nglitch:-self.nglitch]
        delta_F2 = np.zeros(len(delta_F0s))
        delta_phi = np.zeros(len(delta_F0s))
        theta = [0, F0, F1, F2]
        delta_thetas = np.atleast_2d(
                np.array([delta_phi, delta_F0s, delta_F1s, delta_F2]).T)

365
366
        thetas = self.calculate_thetas(theta, delta_thetas, tboundaries,
                                       theta0_idx=self.theta0_idx)
367
368

        twoFSum = 0
369
        for i, theta_i_at_tref in enumerate(thetas):
370
371
372
            ts, te = tboundaries[i], tboundaries[i+1]

            twoFVal = self.run_computefstatistic_single_point(
373
374
                ts, te, theta_i_at_tref[1], theta_i_at_tref[2],
                theta_i_at_tref[3], Alpha, Delta)
375
376
377
378
379
380
            twoFSum += twoFVal

        return twoFSum

    def compute_glitch_fstat_single(self, F0, F1, F2, Alpha, Delta, delta_F0,
                                    delta_F1, tglitch):
381
382
383
384
        """ Returns the semi-coherent glitch summed twoF for nglitch=1

        Note: used for testing
        """
385
386
387
388
389
390
391
392
393
394
395

        theta = [F0, F1, F2]
        delta_theta = [delta_F0, delta_F1, 0]
        tref = self.tref

        theta_at_glitch = self.shift_coefficients(theta, tglitch - tref)
        theta_post_glitch_at_glitch = theta_at_glitch + delta_theta
        theta_post_glitch = self.shift_coefficients(
            theta_post_glitch_at_glitch, tref - tglitch)

        twoFsegA = self.run_computefstatistic_single_point(
Gregory Ashton's avatar
Gregory Ashton committed
396
            self.tstart, tglitch, theta[0], theta[1], theta[2], Alpha,
397
398
399
400
401
402
            Delta)

        if tglitch == self.tend:
            return twoFsegA

        twoFsegB = self.run_computefstatistic_single_point(
Gregory Ashton's avatar
Gregory Ashton committed
403
            tglitch, self.tend, theta_post_glitch[0],
404
405
406
407
408
409
            theta_post_glitch[1], theta_post_glitch[2], Alpha,
            Delta)

        return twoFsegA + twoFsegB


Gregory Ashton's avatar
Gregory Ashton committed
410
411
class MCMCSearch(BaseSearchClass):
    """ MCMC search using ComputeFstat"""
412
    @initializer
413
414
    def __init__(self, label, outdir, sftlabel, sftdir, theta_prior, tref,
                 tstart, tend, nsteps=[100, 100, 100], nwalkers=100, ntemps=1,
415
416
                 log10temperature_min=-5, theta_initial=None, scatter_val=1e-4,
                 binary=False, minCoverFreq=None, maxCoverFreq=None,
Gregory Ashton's avatar
Gregory Ashton committed
417
                 detector=None, earth_ephem=None, sun_ephem=None):
418
419
420
421
422
423
        """
        Parameters
        label, outdir: str
            A label and directory to read/write data from/to
        sftlabel, sftdir: str
            A label and directory in which to find the relevant sft file
424
        theta_prior: dict
425
426
427
428
            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.
429
430
431
432
        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.
433
434
435
436
437
438
439
        tref, tstart, tend: int
            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].
440
441
442
443
444
445
446
447
448
449
450
        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
        detector: str
            Two character reference to the data to use, specify None for no
            contraint.
451
452
453
454
455
456
457
458
459
460
        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
461
462
463
        logging.info(
            'Set-up MCMC search for model {} on data {}'.format(
                self.label, self.sftlabel))
464
465
466
        if os.path.isdir(outdir) is False:
            os.mkdir(outdir)
        self.pickle_path = '{}/{}_saved_data.p'.format(self.outdir, self.label)
Gregory Ashton's avatar
Gregory Ashton committed
467
468
        self.theta_prior['tstart'] = self.tstart
        self.theta_prior['tend'] = self.tend
469
470
        self.unpack_input_theta()
        self.ndim = len(self.theta_keys)
471
        self.betas = np.logspace(0, self.log10temperature_min, self.ntemps)
472
        self.sft_filepath = self.sftdir+'/*_'+self.sftlabel+"*sft"
473

474
475
476
477
478
479
480
481
482
483
484
485
        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()

    def inititate_search_object(self):
        logging.info('Setting up search object')
Gregory Ashton's avatar
Gregory Ashton committed
486
487
488
        self.search = ComputeFstat(
            tref=self.tref, sftlabel=self.sftlabel,
            sftdir=self.sftdir, minCoverFreq=self.minCoverFreq,
489
            maxCoverFreq=self.maxCoverFreq, earth_ephem=self.earth_ephem,
490
            sun_ephem=self.sun_ephem, detector=self.detector, transient=False)
491
492

    def logp(self, theta_vals, theta_prior, theta_keys, search):
Gregory Ashton's avatar
Gregory Ashton committed
493
        H = [self.generic_lnprior(**theta_prior[key])(p) for p, key in
494
495
496
497
498
499
             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]
Gregory Ashton's avatar
Gregory Ashton committed
500
        FS = search.run_computefstatistic_single_point(*self.fixed_theta)
501
502
503
        return FS

    def unpack_input_theta(self):
Gregory Ashton's avatar
Gregory Ashton committed
504
505
        full_theta_keys = ['tstart', 'tend', 'F0', 'F1', 'F2', 'Alpha',
                           'Delta']
506
507
508
        if self.binary:
            full_theta_keys += [
                'asini', 'period', 'ecc', 'tp', 'argp']
509
510
        full_theta_keys_copy = copy.copy(full_theta_keys)

Gregory Ashton's avatar
Gregory Ashton committed
511
512
        full_theta_symbols = ['_', '_', '$f$', '$\dot{f}$', '$\ddot{f}$',
                              r'$\alpha$', r'$\delta$']
513
514
515
516
        if self.binary:
            full_theta_symbols += [
                'asini', 'period', 'period', 'ecc', 'tp', 'argp']

517
518
        self.theta_keys = []
        fixed_theta_dict = {}
519
        for key, val in self.theta_prior.iteritems():
520
521
            if type(val) is dict:
                fixed_theta_dict[key] = 0
Gregory Ashton's avatar
Gregory Ashton committed
522
                self.theta_keys.append(key)
523
524
525
526
527
528
            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
529
            full_theta_keys_copy.pop(full_theta_keys_copy.index(key))
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573

        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):
        initial_priors = np.array([
            self.logp(p, self.theta_prior, self.theta_keys, self.search)
            for p in p0[0]])
        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))

    def run(self):

        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),
            loglargs=(self.search,), betas=self.betas)

Gregory Ashton's avatar
Gregory Ashton committed
574
575
        p0 = self.generate_initial_p0()
        p0 = self.apply_corrections_to_p0(p0)
576
577
578
579
580
581
582
        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(
                j, ninit_steps, n))
            sampler.run_mcmc(p0, n)
583
584
            logging.info("Mean acceptance fraction: {0:.3f}"
                         .format(np.mean(sampler.acceptance_fraction)))
Gregory Ashton's avatar
Gregory Ashton committed
585
            fig, axes = self.plot_walkers(sampler, symbols=self.theta_symbols)
586
587
588
            fig.savefig('{}/{}_init_{}_walkers.png'.format(
                self.outdir, self.label, j))

589
            p0 = self.get_new_p0(sampler)
Gregory Ashton's avatar
Gregory Ashton committed
590
            p0 = self.apply_corrections_to_p0(p0)
591
592
593
594
595
596
597
598
            self.check_initial_points(p0)
            sampler.reset()

        nburn = self.nsteps[-2]
        nprod = self.nsteps[-1]
        logging.info('Running final burn and prod with {} steps'.format(
            nburn+nprod))
        sampler.run_mcmc(p0, nburn+nprod)
599
600
        logging.info("Mean acceptance fraction: {0:.3f}"
                     .format(np.mean(sampler.acceptance_fraction)))
601

Gregory Ashton's avatar
Gregory Ashton committed
602
        fig, axes = self.plot_walkers(sampler, symbols=self.theta_symbols)
603
604
605
606
607
608
609
610
611
612
613
        fig.savefig('{}/{}_walkers.png'.format(self.outdir, self.label))

        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)

614
    def plot_corner(self, figsize=(7, 7),  tglitch_ratio=False,
615
616
617
                    add_prior=False, nstds=None, label_offset=0.4, **kwargs):

        fig, axes = plt.subplots(self.ndim, self.ndim,
618
                                 figsize=figsize)
619
620
621
622
623
624

        samples_plt = copy.copy(self.samples)
        theta_symbols_plt = copy.copy(self.theta_symbols)
        theta_symbols_plt = [s.replace('_{glitch}', r'_\textrm{glitch}') for s
                             in theta_symbols_plt]

625
626
627
628
629
630
631
        if tglitch_ratio:
            for j, k in enumerate(self.theta_keys):
                if k == 'tglitch':
                    s = samples_plt[:, j]
                    samples_plt[:, j] = (s - self.tstart)/(
                                         self.tend - self.tstart)
                    theta_symbols_plt[j] = r'$R_{\textrm{glitch}}$'
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678

        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))

    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
679
            prior = self.generic_lnprior(**self.theta_prior[key])
680
681
682
683
684
685
            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)

Gregory Ashton's avatar
Gregory Ashton committed
686
    def generic_lnprior(self, **kwargs):
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
        """ 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):
            if x < 0:
                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'])
        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
736
    def generate_rv(self, **kwargs):
737
738
739
740
741
742
743
744
745
746
747
748
749
750
        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']))
        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
751
752
    def plot_walkers(self, sampler, symbols=None, alpha=0.4, color="k", temp=0,
                     start=None, stop=None, draw_vline=None):
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
        """ Plot all the chains from a sampler """

        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, :, :, :]

        with plt.style.context(('classic')):
            fig, axes = plt.subplots(ndim, 1, sharex=True, figsize=(8, 4*ndim))

            if ndim > 1:
                for i in range(ndim):
773
                    axes[i].ticklabel_format(useOffset=False, axis='y')
774
775
                    cs = chain[:, start:stop, i].T
                    axes[i].plot(cs, color="k", alpha=alpha)
776
777
778
779
                    if symbols:
                        axes[i].set_ylabel(symbols[i])
                    if draw_vline is not None:
                        axes[i].axvline(draw_vline, lw=2, ls="--")
780
781
782
783
784

            else:
                cs = chain[:, start:stop, 0].T
                axes.plot(cs, color='k', alpha=alpha)
                axes.ticklabel_format(useOffset=False, axis='y')
785
786
787

        return fig, axes

Gregory Ashton's avatar
Gregory Ashton committed
788
789
790
791
792
    def apply_corrections_to_p0(self, p0):
        """ Apply any correction to the initial p0 values """
        return p0

    def generate_scattered_p0(self, p):
793
        """ Generate a set of p0s scattered about p """
Gregory Ashton's avatar
Gregory Ashton committed
794
        p0 = [[p + self.scatter_val * p * np.random.randn(self.ndim)
795
796
797
798
               for i in xrange(self.nwalkers)]
              for j in xrange(self.ntemps)]
        return p0

Gregory Ashton's avatar
Gregory Ashton committed
799
    def generate_initial_p0(self):
800
801
802
        """ Generate a set of init vals for the walkers """

        if type(self.theta_initial) == dict:
Gregory Ashton's avatar
Gregory Ashton committed
803
            p0 = [[[self.generate_rv(**self.theta_initial[key])
804
805
806
807
                    for key in self.theta_keys]
                   for i in range(self.nwalkers)]
                  for j in range(self.ntemps)]
        elif self.theta_initial is None:
Gregory Ashton's avatar
Gregory Ashton committed
808
            p0 = [[[self.generate_rv(**self.theta_prior[key])
809
810
811
812
                    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
813
            p0 = self.generate_scattered_p0(self.theta_initial)
814
815
816
817
818
        else:
            raise ValueError('theta_initial not understood')

        return p0

819
    def get_new_p0(self, sampler):
820
821
822
823
824
825
826
827
828
829
830
831
832
833
        """ 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`.

        """
        if sampler.chain[:, :, -1, :].shape[0] == 1:
            ntemps_temp = 1
        else:
            ntemps_temp = self.ntemps
        pF = sampler.chain[:, :, -1, :].reshape(
            ntemps_temp, self.nwalkers, self.ndim)[0, :, :]
        lnp = sampler.lnprobability[:, :, -1].reshape(
            self.ntemps, self.nwalkers)[0, :]
834
835

        # General warnings about the state of lnp
836
        if any(np.isnan(lnp)):
837
838
839
840
841
842
843
844
845
846
847
            logging.warning(
                "Of {} lnprobs {} are nan".format(
                    len(lnp), np.sum(np.isnan(lnp))))
        if any(np.isposinf(lnp)):
            logging.warning(
                "Of {} lnprobs {} are +np.inf".format(
                    len(lnp), np.sum(np.isposinf(lnp))))
        if any(np.isneginf(lnp)):
            logging.warning(
                "Of {} lnprobs {} are -np.inf".format(
                    len(lnp), np.sum(np.isneginf(lnp))))
848

849
850
851
        lnp_finite = copy.copy(lnp)
        lnp_finite[np.isinf(lnp)] = np.nan
        p = pF[np.nanargmax(lnp_finite)]
852
        p0 = self.generate_scattered_p0(p)
853
854
855
856
857
858

        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
859
                 theta_prior=self.theta_prior, scatter_val=self.scatter_val,
860
861
                 log10temperature_min=self.log10temperature_min,
                 theta0_idx=self.theta0_idx)
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
        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_list_of_matching_sfts(self):
        matches = glob.glob(self.sft_filepath)
        if len(matches) > 0:
            return matches
        else:
            raise IOError('No sfts found matching {}'.format(
                self.sft_filepath))

    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):
892
893
894
895
        if args.use_old_data:
            logging.info("Forcing use of old data")
            return True

896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
        if os.path.isfile(self.pickle_path) is False:
            logging.info('No pickled data found')
            return False

        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

        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:
920
                raise ValueError('Keys {} not in old dictionary'.format(key))
921
922
923
924
925
926
927
928
929

        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':
930
                        logging.info("    {} : {} -> {}".format(*key))
931
                    else:
932
                        logging.info("    " + key[0])
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
                else:
                    logging.info(key)
            return False

    def get_max_twoF(self, threshold=0.05):
        """ Returns the max 2F sample and the corresponding 2F value

        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])
        maxtwoF = self.lnlikes[jmax]
954
        d = OrderedDict()
955
956
957
958

        lnl_finite = copy.copy(self.lnlikes)
        lnl_finite[idxs] = np.nan
        close_idxs = abs((maxtwoF - lnl_finite) / maxtwoF) < threshold
959
960
961
        for i, k in enumerate(self.theta_keys):
            ng = 1
            while k in d:
962
                k = k + '_{}'.format(ng)
963
964
965
966
967
968
969
970
            d[k] = self.samples[jmax][i]

            s = self.samples[:, i][close_idxs]
            d[k + '_std'] = np.std(s)
        return d, maxtwoF

    def get_median_stds(self):
        """ Returns a dict of the median and std of all production samples """
971
        d = OrderedDict()
972
        for s, k in zip(self.samples.T, self.theta_keys):
973
974
975
            ng = 1
            while k in d:
                k = k + '_{}'.format(ng)
976
977
978
979
980
981
982
983
            d[k] = np.median(s)
            d[k+'_std'] = np.std(s)
        return d

    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))
984
985
986
987
988
989
990

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

        filename = '{}/{}.par'.format(self.outdir, self.label)
        with open(filename, 'w+') as f:
            f.write('MaxtwoF = {}\n'.format(max_twoF))
991
            f.write('theta0_index = {}\n'.format(self.theta0_idx))
992
            if method == 'med':
993
994
                for key, val in median_std_d.iteritems():
                    f.write('{} = {:1.16e}\n'.format(key, val))
995
            if method == 'twoFmax':
996
997
998
999
1000
                for key, val in max_twoF_d.iteritems():
                    f.write('{} = {:1.16e}\n'.format(key, val))

    def print_summary(self):
        d, max_twoF = self.get_max_twoF()
1001
        median_std_d = self.get_median_stds()
1002
        print('Max twoF: {}'.format(max_twoF))
1003
        print('theta0 index: {}'.format(self.theta0_idx))
1004
        for k in np.sort(median_std_d.keys()):
1005
1006
            if 'std' not in k:
                print('{:10s} = {:1.9e} +/- {:1.9e}'.format(
1007
                    k, median_std_d[k], median_std_d[k+'_std']))
1008
1009


Gregory Ashton's avatar
Gregory Ashton committed
1010
1011
1012
1013
class MCMCGlitchSearch(MCMCSearch):
    """ MCMC search using the SemiCoherentGlitchSearch """
    @initializer
    def __init__(self, label, outdir, sftlabel, sftdir, theta_prior, tref,
1014
1015
                 tstart, tend, nglitch=1, nsteps=[100, 100, 100], nwalkers=100,
                 ntemps=1, log10temperature_min=-5, theta_initial=None,
1016
1017
                 scatter_val=1e-4, dtglitchmin=1*86400, theta0_idx=0,
                 detector=None,
1018
                 minCoverFreq=None, maxCoverFreq=None, earth_ephem=None,
Gregory Ashton's avatar
Gregory Ashton committed
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
                 sun_ephem=None):
        """
        Parameters
        label, outdir: str
            A label and directory to read/write data from/to
        sftlabel, sftdir: str
            A label and directory in which to find the relevant sft file
        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
            scattered by scatter_val, or  None in which case the prior is used.
1035
1036
1037
1038
        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
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
        nglitch: int
            The number of glitches to allow
        tref, tstart, tend: int
            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
1051
1052
1053
1054
1055
1056
        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).
1057
1058
1059
1060
        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)
1061
1062
1063
        detector: str
            Two character reference to the data to use, specify None for no
            contraint.
Gregory Ashton's avatar
Gregory Ashton committed
1064
1065
1066
1067
1068
1069
1070
1071
1072
1073
1074
1075
1076
1077
1078
1079
1080
1081
        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

        """

        logging.info(('Set-up MCMC glitch search with {} glitches for model {}'
                      ' on data {}').format(self.nglitch, self.label,
                                            self.sftlabel))
        if os.path.isdir(outdir) is False:
            os.mkdir(outdir)
        self.pickle_path = '{}/{}_saved_data.p'.format(self.outdir, self.label)
        self.unpack_input_theta()
        self.ndim = len(self.theta_keys)
1082
        self.betas = np.logspace(0, self.log10temperature_min, self.ntemps)
Gregory Ashton's avatar
Gregory Ashton committed
1083
1084
1085
1086
1087
1088
1089
1090
1091
1092
1093
1094
1095
1096
1097
1098
1099
1100
1101
        self.sft_filepath = self.sftdir+'/*_'+self.sftlabel+"*sft"
        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()

    def inititate_search_object(self):
        logging.info('Setting up search object')
        self.search = SemiCoherentGlitchSearch(
            label=self.label, outdir=self.outdir, sftlabel=self.sftlabel,
            sftdir=self.sftdir, tref=self.tref, tstart=self.tstart,
            tend=self.tend, minCoverFreq=self.minCoverFreq,
            maxCoverFreq=self.maxCoverFreq, earth_ephem=self.earth_ephem,
            sun_ephem=self.sun_ephem, detector=self.detector,
1102
            nglitch=self.nglitch, theta0_idx=self.theta0_idx)
Gregory Ashton's avatar
Gregory Ashton committed
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
1136
1137
1138
1139
1140
1141
1142
1143
1144
1145
1146
1147
1148
1149
1150
1151
1152
1153
1154
1155
1156
1157
1158
1159
1160
1161
1162
1163
1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
1174
1175
1176
1177
1178
1179
1180
1181
1182
1183
1184

    def logp(self, theta_vals, theta_prior, theta_keys, search):
        if self.nglitch > 1:
            ts = [self.tstart] + theta_vals[-self.nglitch:] + [self.tend]
            if np.array_equal(ts, np.sort(ts)) is False:
                return -np.inf
            if any(np.diff(ts) < self.dtglitchmin):
                return -np.inf

        H = [self.generic_lnprior(**theta_prior[key])(p) for p, key in
             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]
        FS = search.compute_nglitch_fstat(*self.fixed_theta)
        return FS

    def unpack_input_theta(self):
        glitch_keys = ['delta_F0', 'delta_F1', 'tglitch']
        full_glitch_keys = list(np.array(
            [[gk]*self.nglitch for gk in glitch_keys]).flatten())
        full_theta_keys = ['F0', 'F1', 'F2', 'Alpha', 'Delta']+full_glitch_keys
        full_theta_keys_copy = copy.copy(full_theta_keys)

        glitch_symbols = ['$\delta f$', '$\delta \dot{f}$', r'$t_{glitch}$']
        full_glitch_symbols = list(np.array(
            [[gs]*self.nglitch for gs in glitch_symbols]).flatten())
        full_theta_symbols = (['$f$', '$\dot{f}$', '$\ddot{f}$', r'$\alpha$',
                               r'$\delta$'] + full_glitch_symbols)
        self.theta_keys = []
        fixed_theta_dict = {}
        for key, val in self.theta_prior.iteritems():
            if type(val) is dict:
                fixed_theta_dict[key] = 0
                if key in glitch_keys:
                    for i in range(self.nglitch):
                        self.theta_keys.append(key)
                else:
                    self.theta_keys.append(key)
            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))
            if key in glitch_keys:
                for i in range(self.nglitch):
                    full_theta_keys_copy.pop(full_theta_keys_copy.index(key))
            else:
                full_theta_keys_copy.pop(full_theta_keys_copy.index(key))

        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]

        # Correct for number of glitches in the idxs
        self.theta_idxs = np.array(self.theta_idxs)
        while np.sum(self.theta_idxs[:-1] == self.theta_idxs[1:]) > 0:
            for i, idx in enumerate(self.theta_idxs):
                if idx in self.theta_idxs[:i]:
                    self.theta_idxs[i] += 1

    def apply_corrections_to_p0(self, p0):
        p0 = np.array(p0)
        if self.nglitch > 1:
            p0[:, :, -self.nglitch:] = np.sort(p0[:, :, -self.nglitch:],
                                               axis=2)
        return p0


Gregory Ashton's avatar
Gregory Ashton committed
1185
1186
class GridSearch(BaseSearchClass):
    """ Gridded search using ComputeFstat """
1187
1188
    @initializer
    def __init__(self, label, outdir, sftlabel=None, sftdir=None, F0s=[0],
Gregory Ashton's avatar
Gregory Ashton committed
1189
1190
1191
1192
                 F1s=[0], F2s=[0], Alphas=[0], Deltas=[0], tref=None,
                 tstart=None, tend=None, minCoverFreq=None, maxCoverFreq=None,
                 write_after=1000, earth_ephem=None, sun_ephem=None,
                 detector=None):
1193
1194
1195
1196
1197
1198
1199
1200
1201
1202
1203
1204
1205
1206
1207
1208
1209
1210
1211
1212
1213
1214
1215
1216
1217
1218
1219
1220
1221
        """
        Parameters
        label, outdir: str
            A label and directory to read/write data from/to
        sftlabel, sftdir: str
            A label and directory in which to find the relevant sft file
        F0s, F1s, F2s, delta_F0s, delta_F1s, tglitchs, Alphas, Deltas: tuple
            Length 3 tuple describing the grid for each parameter, e.g
            [F0min, F0max, dF0], for a fixed value simply give [F0].
        tref, tstart, tend: int
            GPS seconds of the reference time, start time and end time
        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

        """
        if sftlabel is None:
            self.sftlabel = self.label
        if sftdir is None:
            self.sftdir = self.outdir
        if earth_ephem is None:
            self.earth_ephem = self.earth_ephem_default
        if sun_ephem is None:
            self.sun_ephem = self.sun_ephem_default

Gregory Ashton's avatar
Gregory Ashton committed
1222
1223
1224
1225
1226
        self.search = ComputeFstat(
            tref=self.tref, sftlabel=self.sftlabel,
            sftdir=self.sftdir, minCoverFreq=self.minCoverFreq,
            maxCoverFreq=self.maxCoverFreq, earth_ephem=self.earth_ephem,
            sun_ephem=self.sun_ephem, detector=self.detector, transient=False)
1227
1228
1229
1230

        if os.path.isdir(outdir) is False:
            os.mkdir(outdir)
        self.out_file = '{}/{}_gridFS.txt'.format(self.outdir, self.label)
Gregory Ashton's avatar
Gregory Ashton committed
1231
        self.keys = ['_', '_', 'F0', 'F1', 'F2', 'Alpha', 'Delta']
1232
1233
1234
1235
1236
1237
1238
1239
1240

    def get_array_from_tuple(self, x):
        if len(x) == 1:
            return np.array(x)
        else:
            return np.arange(x[0], x[1]*(1+1e-15), x[2])

    def get_input_data_array(self):
        arrays = []
Gregory Ashton's avatar
Gregory Ashton committed
1241
1242
        for tup in ([self.tstart], [self.tend], self.F0s, self.F1s, self.F2s,
                    self.Alphas, self.Deltas):
1243
1244
1245
1246
1247
1248
1249
1250
1251
1252
1253
1254
1255
1256
1257
1258
1259
1260
1261
1262
1263
1264
1265
1266
1267
1268
1269
1270
1271
1272
1273
1274
1275
1276
1277
1278
            arrays.append(self.get_array_from_tuple(tup))

        input_data = []
        for vals in itertools.product(*arrays):
            input_data.append(vals)

        self.arrays = arrays
        self.input_data = np.array(input_data)

    def check_old_data_is_okay_to_use(self):
        if os.path.isfile(self.out_file) is False:
            logging.info('No old data found, continuing with grid search')
            return False
        data = np.atleast_2d(np.genfromtxt(self.out_file, delimiter=' '))
        if np.all(data[:, 0:-1] == self.input_data):
            logging.info(
                'Old data found with matching input, no search performed')
            return data
        else:
            logging.info(
                'Old data found, input differs, continuing with grid search')
            return False

    def run(self):
        self.get_input_data_array()
        old_data = self.check_old_data_is_okay_to_use()
        if old_data is not False:
            self.data = old_data
            return

        logging.info('Total number of grid points is {}'.format(
            len(self.input_data)))

        counter = 0
        data = []
        for vals in self.input_data:
Gregory Ashton's avatar
Gregory Ashton committed
1279
            FS = self.search.run_computefstatistic_single_point(*vals)
1280
1281
1282
1283
1284
1285
1286
1287
1288
1289
1290
            data.append(list(vals) + [FS])

            if counter > self.write_after:
                np.savetxt(self.out_file, data, delimiter=' ')
                counter = 0
                data = []

        logging.info('Saving data to {}'.format(self.out_file))
        np.savetxt(self.out_file, data, delimiter=' ')
        self.data = np.array(data)

Gregory Ashton's avatar
Gregory Ashton committed
1291
1292
1293
1294
1295
1296
1297
1298
    def plot_1D(self, xkey):
        fig, ax = plt.subplots()
        xidx = self.keys.index(xkey)
        x = np.unique(self.data[:, xidx])
        z = self.data[:, -1]
        plt.plot(x, z)
        fig.savefig('{}/{}_1D.png'.format(self.outdir, self.label))

1299
1300
1301
1302
1303
1304
1305
1306
1307
1308
1309
1310
1311
1312
1313
1314
1315
1316
1317
1318
1319
1320
1321
1322
1323
    def plot_2D(self, xkey, ykey):
        fig, ax = plt.subplots()
        xidx = self.keys.index(xkey)
        yidx = self.keys.index(ykey)
        x = np.unique(self.data[:, xidx])
        y = np.unique(self.data[:, yidx])
        z = self.data[:, -1]

        X, Y = np.meshgrid(x, y)
        Z = z.reshape(X.shape)

        pax = ax.pcolormesh(X, Y, Z, cmap=plt.cm.viridis)
        fig.colorbar(pax)
        ax.set_xlim(x[0], x[-1])
        ax.set_ylim(y[0], y[-1])
        ax.set_xlabel(xkey)
        ax.set_ylabel(ykey)

        fig.tight_layout()
        fig.savefig('{}/{}_2D.png'.format(self.outdir, self.label))

    def get_max_twoF(self):
        twoF = self.data[:, -1]
        return np.max(twoF)

1324

Gregory Ashton's avatar
Gregory Ashton committed
1325
1326
1327
1328
1329
1330
1331
1332
1333
1334
1335
1336
1337
1338
1339
1340
1341
1342
1343
1344
1345
1346
1347
1348
1349
1350
1351
1352
1353
1354
1355
1356
1357
1358
1359
1360
1361
1362
1363
1364
1365
1366
1367
1368
1369
1370
1371
1372
1373
1374
1375
1376
1377
1378
1379
1380
1381
1382
1383
1384
1385
1386
1387
1388
1389
class GridGlitchSearch(GridSearch):
    """ Gridded search using the SemiCoherentGlitchSearch """
    @initializer
    def __init__(self, label, outdir, sftlabel=None, sftdir=None, F0s=[0],
                 F1s=[0], F2s=[0], delta_F0s=[0], delta_F1s=[0], tglitchs=None,
                 Alphas=[0], Deltas=[0], tref=None, tstart=None, tend=None,
                 minCoverFreq=None, maxCoverFreq=None, write_after=1000,
                 earth_ephem=None, sun_ephem=None):
        """
        Parameters
        label, outdir: str
            A label and directory to read/write data from/to
        sftlabel, sftdir: str
            A label and directory in which to find the relevant sft file
        F0s, F1s, F2s, delta_F0s, delta_F1s, tglitchs, Alphas, Deltas: tuple
            Length 3 tuple describing the grid for each parameter, e.g
            [F0min, F0max, dF0], for a fixed value simply give [F0].
        tref, tstart, tend: int
            GPS seconds of the reference time, start time and end time
        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

        """
        if tglitchs is None:
            self.tglitchs = [self.tend]
        if sftlabel is None:
            self.sftlabel = self.label
        if sftdir is None:
            self.sftdir = self.outdir
        if earth_ephem is None:
            self.earth_ephem = self.earth_ephem_default
        if sun_ephem is None:
            self.sun_ephem = self.sun_ephem_default

        self.search = SemiCoherentGlitchSearch(
            label=label, outdir=outdir, sftlabel=sftlabel, sftdir=sftdir,
            tref=tref, tstart=tstart, tend=tend, minCoverFreq=minCoverFreq,
            maxCoverFreq=maxCoverFreq, earth_ephem=self.earth_ephem,
            sun_ephem=self.sun_ephem)

        if os.path.isdir(outdir) is False:
            os.mkdir(outdir)
        self.out_file = '{}/{}_gridFS.txt'.format(self.outdir, self.label)
        self.keys = ['F0', 'F1', 'F2', 'Alpha', 'Delta', 'delta_F0',
                     'delta_F1', 'tglitch']

    def get_input_data_array(self):
        arrays = []
        for tup in (self.F0s, self.F1s, self.F2s, self.Alphas, self.Deltas,
                    self.delta_F0s, self.delta_F1s, self.tglitchs):
            arrays.append(self.get_array_from_tuple(tup))

        input_data = []
        for vals in itertools.product(*arrays):
            input_data.append(vals)

        self.arrays = arrays
        self.input_data = np.array(input_data)


1390
1391
1392
1393
1394
1395
1396
1397
class Writer(BaseSearchClass):
    """ Instance object for generating SFTs containing glitch signals """
    @initializer
    def __init__(self, label='Test', tstart=700000000, duration=100*86400,
                 dtglitch=None,
                 delta_phi=0, delta_F0=0, delta_F1=0, delta_F2=0,
                 tref=None, phi=0, F0=30, F1=1e-10, F2=0, Alpha=5e-3,
                 Delta=6e-2, h0=0.1, cosi=0.0, psi=0.0, Tsft=1800, outdir=".",
Gregory Ashton's avatar
Gregory Ashton committed
1398
                 sqrtSX=1, Band=4, detector='H1'):
1399
1400
1401
1402
1403
1404
1405
1406
1407
1408
1409
1410
1411
1412
1413
1414
1415
1416
1417
1418
1419
1420
1421
1422
1423
1424
1425
1426
1427
1428
1429
1430
1431
1432
1433
1434
1435
1436
1437
1438
1439
1440
1441
1442
1443
1444
1445
1446
1447
1448
1449
1450
1451
1452
1453
1454
1455
1456
1457
1458
1459
1460
1461
1462
1463
1464
1465
1466
1467
1468
1469
1470
1471
1472
1473
1474
1475
1476
1477
1478
1479
1480
1481
1482
1483
1484
1485
1486
1487
1488
1489
1490
1491
1492
1493
1494
1495
1496
1497
1498
1499
1500
1501
1502
1503
1504
1505
1506
1507
1508
1509
1510
1511
1512
1513
1514
1515
1516
1517
1518
1519
1520
1521
1522
1523
1524
1525
1526
1527
1528
1529
1530
1531
1532
1533
1534
1535
1536
1537
1538
1539
1540
1541
1542
1543
1544
1545
1546
1547
1548
1549
1550
1551
1552
1553
1554
1555
1556
1557
1558
1559
1560
1561
1562
1563
1564
1565
1566
1567
1568
1569
1570
1571
1572
1573
1574
1575
1576
1577
1578
1579
1580
1581
1582
1583
1584
1585
1586
1587
1588
1589
1590
1591
1592
1593
1594
1595
1596
1597
1598
1599
1600
1601
1602
1603
1604
1605
1606
1607
1608
1609
1610
1611
1612
1613
1614
1615
1616
1617
1618
1619
        """
        Parameters
        ----------
        label: string
            a human-readable label to be used in naming the output files
        tstart, tend : float
            start and end times (in gps seconds) of the total observation span
        dtglitch: float
            time (in gps seconds) of the glitch after tstart. To create data
            without a glitch, set dtglitch=tend-tstart or leave as None
        delta_phi, delta_F0, delta_F1: float
            instanteneous glitch magnitudes in rad, Hz, and Hz/s respectively
        tref: float or None
            reference time (default is None, which sets the reference time to
            tstart)
        phil, F0, F1, F2, Alpha, Delta, h0, cosi, psi: float
            pre-glitch phase, frequency, sky-position, and signal properties
        Tsft: float
            the sft duration

        see `lalapps_Makefakedata_v5 --help` for help with the other paramaters
        """

        for d in self.delta_phi, self.delta_F0, self.delta_F1, self.delta_F2:
            if np.size(d) == 1:
                d = [d]
        self.tend = self.tstart + self.duration
        if self.dtglitch is None or self.dtglitch == self.duration:
            self.tbounds = [self.tstart, self.tend]
        elif np.size(self.dtglitch) == 1:
            self.tbounds = [self.tstart, self.tstart+self.dtglitch, self.tend]
        else:
            self.tglitch = self.tstart + np.array(self.dtglitch)
            self.tbounds = [self.tstart] + list(self.tglitch) + [self.tend]

        if os.path.isdir(self.outdir) is False:
            os.makedirs(self.outdir)
        if self.tref is None:
            self.tref = self.tstart
        self.tend = self.tstart + self.duration
        tbs = np.array(self.tbounds)
        self.durations_days = (tbs[1:] - tbs[:-1]) / 86400
        self.config_file_name = "{}/{}.cff".format(outdir, label)

        self.theta = np.array([phi, F0, F1, F2])
        self.delta_thetas = np.atleast_2d(
                np.array([delta_phi, delta_F0, delta_F1, delta_F2]).T)

        numSFTs = int(float(self.duration) / self.Tsft)
        self.sft_filename = lalpulsar.OfficialSFTFilename(
            'H', '1', numSFTs, self.Tsft, self.tstart, self.duration,
            self.label)
        self.sft_filepath = '{}/{}'.format(self.outdir, self.sft_filename)
        self.calculate_fmin_Band()

    def make_data(self):
        ''' A convienience wrapper to generate a cff file then sfts '''
        self.make_cff()
        self.run_makefakedata()

    def get_single_config_line(self, i, Alpha, Delta, h0, cosi, psi, phi, F0,
                               F1, F2, tref, tstart, duration_days):
        template = (
"""[TS{}]
Alpha = {:1.18e}
Delta = {:1.18e}
h0 = {:1.18e}
cosi = {:1.18e}
psi = {:1.18e}
phi0 = {:1.18e}
Freq = {:1.18e}
f1dot = {:1.18e}
f2dot = {:1.18e}
refTime = {:10.6f}
transientWindowType=rect
transientStartTime={:10.3f}
transientTauDays={:1.3f}\n""")
        return template.format(i, Alpha, Delta, h0, cosi, psi, phi, F0, F1,
                               F2, tref, tstart, duration_days)

    def make_cff(self):
        """
        Generates an .cff file for a 'glitching' signal

        """

        thetas = self.calculate_thetas(self.theta, self.delta_thetas,
                                       self.tbounds)

        content = ''
        for i, (t, d, ts) in enumerate(zip(thetas, self.durations_days,
                                           self.tbounds[:-1])):
            line = self.get_single_config_line(
                i, self.Alpha, self.Delta, self.h0, self.cosi, self.psi,
                t[0], t[1], t[2], t[3], self.tref, ts, d)

            content += line

        if self.check_if_cff_file_needs_rewritting(content):
            config_file = open(self.config_file_name, "w+")
            config_file.write(content)
            config_file.close()

    def calculate_fmin_Band(self):
        self.fmin = self.F0 - .5 * self.Band

    def check_cached_data_okay_to_use(self, cl):
        """ Check if cached data exists and, if it does, if it can be used """

        getmtime = os.path.getmtime

        if os.path.isfile(self.sft_filepath) is False:
            logging.info('No SFT file matching {} found'.format(
                self.sft_filepath))
            return False
        else:
            logging.info('Matching SFT file found')

        if getmtime(self.sft_filepath) < getmtime(self.config_file_name):
            logging.info(
                ('The config file {} has been modified since the sft file {} '
                 + 'was created').format(
                    self.config_file_name, self.sft_filepath))
            return False

        logging.info(
            'The config file {} is older than the sft file {}'.format(
                self.config_file_name, self.sft_filepath))
        logging.info('Checking contents of cff file')
        logging.info('Execute: {}'.format(
            'lalapps_SFTdumpheader {} | head -n 20'.format(self.sft_filepath)))
        output = subprocess.check_output(
            'lalapps_SFTdumpheader {} | head -n 20'.format(self.sft_filepath),
            shell=True)
        calls = [line for line in output.split('\n') if line[:3] == 'lal']
        if calls[0] == cl:
            logging.info('Contents matched, use old sft file')
            return True
        else:
            logging.info('Contents unmatched, create new sft file')
            return False

    def check_if_cff_file_needs_rewritting(self, content):
        """ Check if the .cff file has changed

        Returns True if the file should be overwritten - where possible avoid
        overwriting to allow cached data to be used
        """
        if os.path.isfile(self.config_file_name) is False:
            logging.info('No config file {} found'.format(
                self.config_file_name))
            return True
        else:
            logging.info('Config file {} already exists'.format(
                self.config_file_name))

        with open(self.config_file_name, 'r') as f:
            file_content = f.read()
            if file_content == content:
                logging.info(
                    'File contents match, no update of {} required'.format(
                        self.config_file_name))
                return False
            else:
                logging.info(
                    'File contents unmatched, updating {}'.format(
                        self.config_file_name))
                return True

    def run_makefakedata(self):
        """ Generate the sft data from the configuration file """

        # Remove old data:
        try:
            os.unlink("{}/*{}*.sft".format(self.outdir, self.label))
        except OSError:
            pass

        cl = []
        cl.append('lalapps_Makefakedata_v5')
        cl.append('--outSingleSFT=TRUE')
        cl.append('--outSFTdir="{}"'.format(self.outdir))
        cl.append('--outLabel="{}"'.format(self.label))
        cl.append('--IFOs="{}"'.format(self.detector))
        cl.append('--sqrtSX="{}"'.format(self.sqrtSX))
        cl.append('--startTime={:10.9f}'.format(float(self.tstart)))
        cl.append('--duration={}'.format(int(self.duration)))
        cl.append('--fmin={}'.format(int(self.fmin)))
        cl.append('--Band={}'.format(self.Band))
        cl.append('--Tsft={}'.format(self.Tsft))
        cl.append('--injectionSources="./{}"'.format(self.config_file_name))

        cl = " ".join(cl)

        if self.check_cached_data_okay_to_use(cl) is False:
            logging.info("Executing: " + cl)
            os.system(cl)
            os.system('\n')

    def predict_fstat(self):
        """ Wrapper to lalapps_PredictFstat """
        c_l = []
        c_l.append("lalapps_PredictFstat")
        c_l.append("--h0={}".format(self.h0))
        c_l.append("--cosi={}".format(self.cosi))
        c_l.append("--psi={}".format(self.psi))
        c_l.append("--Alpha={}".format(self.Alpha))
        c_l.append("--Delta={}".format(self.Delta))
        c_l.append("--Freq={}".format(self.F0))

        c_l.append("--DataFiles='{}'".format(
            self.outdir+"/*SFT_"+self.label+"*sft"))
        c_l.append("--assumeSqrtSX={}".format(self.sqrtSX))

        c_l.append("--minStartTime={}".format(self.tstart))
        c_l.append("--maxStartTime={}".format(self.tend))

        logging.info("Executing: " + " ".join(c_l) + "\n")
        output = subprocess.check_output(" ".join(c_l), shell=True)
        twoF = float(output.split('\n')[-2])
        return float(twoF)