correlated_fields.py 23.7 KB
Newer Older
Martin Reinecke's avatar
Martin Reinecke committed
1
2
3
4
5
6
7
8
9
10
11
12
13
# This program is free software: you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation, either version 3 of the License, or
# (at your option) any later version.
#
# This program is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with this program.  If not, see <http://www.gnu.org/licenses/>.
#
14
# Copyright(C) 2013-2019 Max-Planck-Society
Philipp Arras's avatar
Philipp Arras committed
15
# Authors: Philipp Frank, Philipp Arras, Philipp Haim
Martin Reinecke's avatar
Martin Reinecke committed
16
#
17
# NIFTy is being developed at the Max-Planck-Institut fuer Astrophysik.
Philipp Arras's avatar
Philipp Arras committed
18

Philipp Arras's avatar
Philipp Arras committed
19
import numpy as np
20

21
from ..logger import logger
Philipp Arras's avatar
Philipp Arras committed
22
from ..domain_tuple import DomainTuple
Philipp Arras's avatar
Philipp Arras committed
23
24
from ..domains.power_space import PowerSpace
from ..domains.unstructured_domain import UnstructuredDomain
Philipp Arras's avatar
Philipp Arras committed
25
26
from ..field import Field
from ..multi_field import MultiField
Philipp Arras's avatar
Philipp Arras committed
27
from ..operators.adder import Adder
28
from ..operators.contraction_operator import ContractionOperator
Philipp Arras's avatar
Philipp Arras committed
29
from ..operators.diagonal_operator import DiagonalOperator
Philipp Arras's avatar
Philipp Arras committed
30
from ..operators.distributors import PowerDistributor
Philipp Arras's avatar
Philipp Arras committed
31
from ..operators.endomorphic_operator import EndomorphicOperator
Martin Reinecke's avatar
Martin Reinecke committed
32
from ..operators.harmonic_operators import HarmonicTransformOperator
Philipp Arras's avatar
Philipp Arras committed
33
from ..operators.linear_operator import LinearOperator
Philipp Arras's avatar
Philipp Arras committed
34
from ..operators.operator import Operator
Philipp Arras's avatar
Philipp Arras committed
35
from ..operators.simple_linear_operators import ducktape
36
from ..probing import StatCalculator
Philipp Frank's avatar
cleanup    
Philipp Frank committed
37
from ..sugar import from_global_data, full, makeDomain
38

39

Philipp Haim's avatar
Philipp Haim committed
40
41
42
def _reshaper(x, N):
    x = np.asfarray(x)
    if x.shape in [(), (1,)]:
Philipp Haim's avatar
Philipp Haim committed
43
        return np.full(N, x) if N != 0 else x.reshape(())
Philipp Haim's avatar
Philipp Haim committed
44
45
    elif x.shape == (N,):
        return x
46
47
48
    else:
        raise TypeError("Shape of parameters cannot be interpreted")

49

Martin Reinecke's avatar
Martin Reinecke committed
50
def _lognormal_moments(mean, sig, N=0):
Philipp Haim's avatar
Philipp Haim committed
51
52
53
54
    if N == 0:
        mean, sig = np.asfarray(mean), np.asfarray(sig)
    else:
        mean, sig = (_reshaper(param, N) for param in (mean, sig))
55
56
57
58
59
    if not np.all(mean > 0):
        raise ValueError(f"mean must be greater 0; got {mean!r}")
    if not np.all(sig > 0):
        raise ValueError(f"sig must be greater 0; got {sig!r}")

Philipp Arras's avatar
Philipp Arras committed
60
61
    logsig = np.sqrt(np.log((sig/mean)**2 + 1))
    logmean = np.log(mean) - logsig**2/2
62
    return logmean, logsig
Philipp Arras's avatar
Philipp Arras committed
63
64


Martin Reinecke's avatar
Martin Reinecke committed
65
def _normal(mean, sig, key, N=0):
Philipp Haim's avatar
Philipp Haim committed
66
    if N == 0:
Philipp Haim's avatar
Philipp Haim committed
67
        domain = DomainTuple.scalar_domain()
Philipp Haim's avatar
Philipp Haim committed
68
        mean, sig = np.asfarray(mean), np.asfarray(sig)
Philipp Haim's avatar
Philipp Haim committed
69
70
    else:
        domain = UnstructuredDomain(N)
Philipp Haim's avatar
Philipp Haim committed
71
        mean, sig = (_reshaper(param, N) for param in (mean, sig))
Philipp Arras's avatar
Philipp Arras committed
72
73
    return Adder(from_global_data(domain, mean)) @ (DiagonalOperator(
        from_global_data(domain, sig)) @ ducktape(domain, None, key))
Philipp Arras's avatar
Philipp Arras committed
74
75


Philipp Arras's avatar
Philipp Arras committed
76
def _log_k_lengths(pspace):
Philipp Arras's avatar
Philipp Arras committed
77
    """Log(k_lengths) without zeromode"""
Philipp Arras's avatar
Philipp Arras committed
78
79
80
    return np.log(pspace.k_lengths[1:])


Philipp Arras's avatar
Philipp Arras committed
81
def _relative_log_k_lengths(power_space):
Philipp Arras's avatar
Philipp Arras committed
82
83
    """Log-distance to first bin
    logkl.shape==power_space.shape, logkl[0]=logkl[1]=0"""
Philipp Arras's avatar
Philipp Arras committed
84
85
86
87
88
89
    power_space = DomainTuple.make(power_space)
    assert isinstance(power_space[0], PowerSpace)
    assert len(power_space) == 1
    logkl = _log_k_lengths(power_space[0])
    assert logkl.shape[0] == power_space[0].shape[0] - 1
    logkl -= logkl[0]
Philipp Arras's avatar
Philipp Arras committed
90
    return np.insert(logkl, 0, 0)
Philipp Arras's avatar
Philipp Arras committed
91
92


Philipp Arras's avatar
Philipp Arras committed
93
def _log_vol(power_space):
94
    power_space = makeDomain(power_space)
Philipp Arras's avatar
Philipp Arras committed
95
96
97
98
99
    assert isinstance(power_space[0], PowerSpace)
    logk_lengths = _log_k_lengths(power_space[0])
    return logk_lengths[1:] - logk_lengths[:-1]


Philipp Haim's avatar
Philipp Haim committed
100
def _total_fluctuation_realized(samples):
101
102
    res = 0.
    for s in samples:
Philipp Haim's avatar
Fixes    
Philipp Haim committed
103
        res = res + (s - s.mean())**2
Philipp Haim's avatar
Philipp Haim committed
104
    return np.sqrt((res/len(samples)).mean())
105
106


Philipp Arras's avatar
Philipp Arras committed
107
class _LognormalMomentMatching(Operator):
Philipp Haim's avatar
Philipp Haim committed
108
    def __init__(self, mean, sig, key, N_copies):
Philipp Arras's avatar
Philipp Arras committed
109
        key = str(key)
Philipp Haim's avatar
Philipp Haim committed
110
        logmean, logsig = _lognormal_moments(mean, sig, N_copies)
Philipp Arras's avatar
Philipp Arras committed
111
112
        self._mean = mean
        self._sig = sig
Philipp Haim's avatar
Philipp Haim committed
113
        op = _normal(logmean, logsig, key, N_copies).exp()
Philipp Arras's avatar
Philipp Arras committed
114
115
116
117
118
119
120
121
122
123
        self._domain, self._target = op.domain, op.target
        self.apply = op.apply

    @property
    def mean(self):
        return self._mean

    @property
    def std(self):
        return self._sig
Philipp Arras's avatar
Philipp Arras committed
124
125


Philipp Frank's avatar
Philipp Frank committed
126
class _SlopeRemover(EndomorphicOperator):
Martin Reinecke's avatar
Martin Reinecke committed
127
    def __init__(self, domain, space=0):
Philipp Frank's avatar
Philipp Frank committed
128
        self._domain = makeDomain(domain)
129
130
        assert isinstance(self._domain[space], PowerSpace)
        logkl = _relative_log_k_lengths(self._domain[space])
131
        self._sc = logkl/float(logkl[-1])
Philipp Arras's avatar
Philipp Arras committed
132

133
        self._space = space
Philipp Haim's avatar
Philipp Haim committed
134
135
136
        axis = self._domain.axes[space][0]
        self._last = (slice(None),)*axis + (-1,) + (None,)
        self._extender = (None,)*(axis) + (slice(None),) + (None,)*(self._domain.axes[-1][-1]-axis)
Philipp Frank's avatar
Philipp Frank committed
137
        self._capability = self.TIMES | self.ADJOINT_TIMES
Philipp Arras's avatar
Philipp Arras committed
138

139
140
    def apply(self, x, mode):
        self._check_input(x, mode)
Philipp Frank's avatar
Philipp Frank committed
141
142
        x = x.to_global_data()
        if mode == self.TIMES:
Philipp Haim's avatar
Philipp Haim committed
143
            res = x - x[self._last]*self._sc[self._extender]
Philipp Frank's avatar
Philipp Frank committed
144
        else:
145
146
            res = x.copy()
            res[self._last] -= (x*self._sc[self._extender]).sum(
Philipp Arras's avatar
Philipp Arras committed
147
                axis=self._space, keepdims=True)
148
        return from_global_data(self._tgt(mode), res)
Philipp Frank's avatar
Philipp Frank committed
149

Philipp Arras's avatar
Philipp Arras committed
150
151

class _TwoLogIntegrations(LinearOperator):
Martin Reinecke's avatar
Martin Reinecke committed
152
    def __init__(self, target, space=0):
Philipp Arras's avatar
Philipp Arras committed
153
        self._target = makeDomain(target)
154
155
156
157
158
        assert isinstance(self.target[space], PowerSpace)
        dom = list(self._target)
        dom[space] = UnstructuredDomain((2, self.target[space].shape[0]-2))
        self._domain = makeDomain(dom)
        self._space = space
159
        self._log_vol = _log_vol(self._target[space])
Philipp Arras's avatar
Philipp Arras committed
160
161
162
163
        self._capability = self.TIMES | self.ADJOINT_TIMES

    def apply(self, x, mode):
        self._check_input(x, mode)
164

Martin Reinecke's avatar
Martin Reinecke committed
165
        # Maybe make class properties
166
167
        axis = self._target.axes[self._space][0]
        sl = (slice(None),)*axis
Philipp Haim's avatar
Fixes    
Philipp Haim committed
168
        extender_sl = (None,)*axis + (slice(None),) + (None,)*(self._target.axes[-1][-1] - axis)
169
170
        first = sl + (0,)
        second = sl + (1,)
Martin Reinecke's avatar
Martin Reinecke committed
171
172
173
        from_third = sl + (slice(2, None),)
        no_border = sl + (slice(1, -1),)
        reverse = sl + (slice(None, None, -1),)
174

Philipp Arras's avatar
Philipp Arras committed
175
        if mode == self.TIMES:
Philipp Haim's avatar
Philipp Haim committed
176
            x = x.to_global_data()
Philipp Arras's avatar
Philipp Arras committed
177
            res = np.empty(self._target.shape)
178
            res[first] = res[second] = 0
Martin Reinecke's avatar
Martin Reinecke committed
179
            res[from_third] = np.cumsum(x[second], axis=axis)
180
            res[from_third] = (res[from_third] + res[no_border])/2*self._log_vol[extender_sl] + x[first]
Martin Reinecke's avatar
Martin Reinecke committed
181
            res[from_third] = np.cumsum(res[from_third], axis=axis)
Philipp Arras's avatar
Philipp Arras committed
182
        else:
Philipp Haim's avatar
Philipp Haim committed
183
            x = x.to_global_data_rw()
Philipp Arras's avatar
Philipp Arras committed
184
            res = np.zeros(self._domain.shape)
Martin Reinecke's avatar
Martin Reinecke committed
185
            x[from_third] = np.cumsum(x[from_third][reverse], axis=axis)[reverse]
186
            res[first] += x[from_third]
187
            x[from_third] *= (self._log_vol/2.)[extender_sl]
188
            x[no_border] += x[from_third]
Martin Reinecke's avatar
Martin Reinecke committed
189
            res[second] += np.cumsum(x[from_third][reverse], axis=axis)[reverse]
190
        return from_global_data(self._tgt(mode), res)
Philipp Arras's avatar
Philipp Arras committed
191
192
193


class _Normalization(Operator):
Martin Reinecke's avatar
Martin Reinecke committed
194
    def __init__(self, domain, space=0):
Philipp Arras's avatar
Philipp Arras committed
195
        self._domain = self._target = makeDomain(domain)
196
        assert isinstance(self._domain[space], PowerSpace)
197
198
199
        hspace = list(self._domain)
        hspace[space] = hspace[space].harmonic_partner
        hspace = makeDomain(hspace)
Philipp Arras's avatar
Philipp Arras committed
200
201
202
        pd = PowerDistributor(hspace,
                              power_space=self._domain[space],
                              space=space)
203
        mode_multiplicity = pd.adjoint(full(pd.target, 1.)).to_global_data_rw()
204
        zero_mode = (slice(None),)*self._domain.axes[space][0] + (0,)
Philipp Haim's avatar
Philipp Haim committed
205
        mode_multiplicity[zero_mode] = 0
Philipp Arras's avatar
Philipp Arras committed
206
207
        self._mode_multiplicity = from_global_data(self._domain,
                                                   mode_multiplicity)
208
        self._specsum = _SpecialSum(self._domain, space)
Philipp Arras's avatar
Philipp Arras committed
209
210
211
212
213
214
215

    def apply(self, x):
        self._check_input(x)
        amp = x.exp()
        spec = (2*x).exp()
        # FIXME This normalizes also the zeromode which is supposed to be left
        # untouched by this operator
216
        return self._specsum(self._mode_multiplicity*spec)**(-0.5)*amp
Philipp Arras's avatar
Philipp Arras committed
217
218
219


class _SpecialSum(EndomorphicOperator):
Martin Reinecke's avatar
Martin Reinecke committed
220
    def __init__(self, domain, space=0):
Philipp Arras's avatar
Philipp Arras committed
221
222
        self._domain = makeDomain(domain)
        self._capability = self.TIMES | self.ADJOINT_TIMES
223
        self._contractor = ContractionOperator(domain, space)
Philipp Arras's avatar
Philipp Arras committed
224
225
226

    def apply(self, x, mode):
        self._check_input(x, mode)
227
        return self._contractor.adjoint(self._contractor(x))
Philipp Arras's avatar
Philipp Arras committed
228
229


Philipp Haim's avatar
Philipp Haim committed
230
class _Distributor(LinearOperator):
Martin Reinecke's avatar
Martin Reinecke committed
231
    def __init__(self, dofdex, domain, target, space=0):
Philipp Haim's avatar
Philipp Haim committed
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
        self._dofdex = dofdex

        self._target = makeDomain(target)
        self._domain = makeDomain(domain)
        self._sl = (slice(None),)*space
        self._capability = self.TIMES | self.ADJOINT_TIMES

    def apply(self, x, mode):
        self._check_input(x, mode)
        x = x.to_global_data()
        if mode == self.TIMES:
            res = x[self._dofdex]
        else:
            res = np.empty(self._tgt(mode).shape)
            res[self._dofdex] = x
        return from_global_data(self._tgt(mode), res)
Martin Reinecke's avatar
Martin Reinecke committed
248

249

250
251
class _Amplitude(Operator):
    def __init__(self, target, fluctuations, flexibility, asperity,
252
                 loglogavgslope, azm, totvol, key, dofdex):
Philipp Arras's avatar
Philipp Arras committed
253
254
255
256
257
258
259
260
261
262
263
        """
        fluctuations > 0
        flexibility > 0
        asperity > 0
        loglogavgslope probably negative
        """
        assert isinstance(fluctuations, Operator)
        assert isinstance(flexibility, Operator)
        assert isinstance(asperity, Operator)
        assert isinstance(loglogavgslope, Operator)

Philipp Haim's avatar
Philipp Haim committed
264
265
        if len(dofdex) > 0:
            N_copies = max(dofdex) + 1
Philipp Haim's avatar
Philipp Haim committed
266
            space = 1
Philipp Frank's avatar
cleanup    
Philipp Frank committed
267
268
            distributed_tgt = makeDomain((UnstructuredDomain(len(dofdex)),
                                          target))
Philipp Haim's avatar
Philipp Haim committed
269
270
271
            target = makeDomain((UnstructuredDomain(N_copies), target))
            Distributor = _Distributor(dofdex, target, distributed_tgt, 0)
        else:
Philipp Haim's avatar
Philipp Haim committed
272
            N_copies = 0
Philipp Haim's avatar
Philipp Haim committed
273
            space = 0
Philipp Haim's avatar
Philipp Haim committed
274
            distributed_tgt = target = makeDomain(target)
Martin Reinecke's avatar
Martin Reinecke committed
275
        azm_expander = ContractionOperator(distributed_tgt, spaces=space).adjoint
Philipp Haim's avatar
Philipp Haim committed
276
        assert isinstance(target[space], PowerSpace)
Martin Reinecke's avatar
Martin Reinecke committed
277

278
        twolog = _TwoLogIntegrations(target, space)
Philipp Arras's avatar
Philipp Arras committed
279
        dom = twolog.domain
280

281
        shp = dom[space].shape
Martin Reinecke's avatar
Martin Reinecke committed
282
283
        expander = ContractionOperator(dom, spaces=space).adjoint
        ps_expander = ContractionOperator(twolog.target, spaces=space).adjoint
Philipp Arras's avatar
Philipp Arras committed
284
285
286

        # Prepare constant fields
        foo = np.zeros(shp)
287
288
        foo[0] = foo[1] = np.sqrt(_log_vol(target[space]))
        vflex = DiagonalOperator(from_global_data(dom[space], foo), dom, space)
Philipp Arras's avatar
Philipp Arras committed
289
290
291

        foo = np.zeros(shp, dtype=np.float64)
        foo[0] += 1
292
        vasp = DiagonalOperator(from_global_data(dom[space], foo), dom, space)
Philipp Arras's avatar
Philipp Arras committed
293
294

        foo = np.ones(shp)
295
296
        foo[0] = _log_vol(target[space])**2/12.
        shift = DiagonalOperator(from_global_data(dom[space], foo), dom, space)
Martin Reinecke's avatar
Martin Reinecke committed
297

298
        vslope = DiagonalOperator(
Martin Reinecke's avatar
Martin Reinecke committed
299
300
301
            from_global_data(target[space],
                             _relative_log_k_lengths(target[space])),
            target, space)
302
303

        foo, bar = [np.zeros(target[space].shape) for _ in range(2)]
Philipp Arras's avatar
Philipp Arras committed
304
        bar[1:] = foo[0] = totvol
Martin Reinecke's avatar
Martin Reinecke committed
305
        vol0, vol1 = [DiagonalOperator(from_global_data(target[space], aa),
Philipp Frank's avatar
cleanup    
Philipp Frank committed
306
                                       target, space) for aa in (foo, bar)]
307

Martin Reinecke's avatar
Martin Reinecke committed
308
        # Prepare fields for Adder
309
        shift, vol0 = [op(full(op.domain, 1)) for op in (shift, vol0)]
Philipp Arras's avatar
Philipp Arras committed
310
311
        # End prepare constant fields

312
313
314
315
        slope = vslope @ ps_expander @ loglogavgslope
        sig_flex = vflex @ expander @ flexibility
        sig_asp = vasp @ expander @ asperity
        sig_fluc = vol1 @ ps_expander @ fluctuations
Philipp Haim's avatar
Philipp Haim committed
316
        sig_fluc = vol1 @ ps_expander @ fluctuations
Philipp Arras's avatar
Philipp Arras committed
317
318

        xi = ducktape(dom, None, key)
Philipp Arras's avatar
Philipp Arras committed
319
        sigma = sig_flex*(Adder(shift) @ sig_asp).sqrt()
320
321
        smooth = _SlopeRemover(target, space) @ twolog @ (sigma*xi)
        op = _Normalization(target, space) @ (slope + smooth)
Philipp Haim's avatar
Philipp Haim committed
322
        if N_copies > 0:
Philipp Haim's avatar
Philipp Haim committed
323
324
            op = Distributor @ op
            sig_fluc = Distributor @ sig_fluc
Philipp Haim's avatar
Philipp Haim committed
325
            op = Adder(Distributor(vol0)) @ (sig_fluc*(azm_expander @ azm.one_over())*op)
Philipp Arras's avatar
Philipp Arras committed
326
327
            self._fluc = (_Distributor(dofdex, fluctuations.target,
                                       distributed_tgt[0]) @ fluctuations)
Philipp Haim's avatar
Philipp Haim committed
328
        else:
Philipp Frank's avatar
cleanup    
Philipp Frank committed
329
            op = Adder(vol0) @ (sig_fluc*(azm_expander @ azm.one_over())*op)
Philipp Frank's avatar
fixup    
Philipp Frank committed
330
            self._fluc = fluctuations
Philipp Arras's avatar
Philipp Arras committed
331

Philipp Arras's avatar
Philipp Arras committed
332
333
        self.apply = op.apply
        self._domain, self._target = op.domain, op.target
334
        self._space = space
Philipp Arras's avatar
Philipp Arras committed
335

Philipp Arras's avatar
Philipp Arras committed
336
337
338
339
    @property
    def fluctuation_amplitude(self):
        return self._fluc

340
341

class CorrelatedFieldMaker:
Philipp Haim's avatar
Philipp Haim committed
342
    def __init__(self, amplitude_offset, prefix, total_N):
343
344
        if not isinstance(amplitude_offset, Operator):
            raise TypeError("amplitude_offset needs to be an operator")
345
        self._a = []
346
        self._position_spaces = []
Philipp Arras's avatar
Formats    
Philipp Arras committed
347

348
349
        self._azm = amplitude_offset
        self._prefix = prefix
Philipp Haim's avatar
Philipp Haim committed
350
        self._total_N = total_N
Philipp Arras's avatar
Formats    
Philipp Arras committed
351

352
    @staticmethod
Philipp Frank's avatar
Philipp Frank committed
353
    def make(offset_amplitude_mean, offset_amplitude_stddev, prefix,
Martin Reinecke's avatar
Martin Reinecke committed
354
355
             total_N=0,
             dofdex=None):
Philipp Frank's avatar
Philipp Frank committed
356
357
        if dofdex is None:
            dofdex = np.full(total_N, 0)
358
359
        elif len(dofdex) != total_N:
            raise ValueError("length of dofdex needs to match total_N")
Philipp Frank's avatar
Philipp Frank committed
360
        N = max(dofdex) + 1 if total_N > 0 else 0
361
362
        zm = _LognormalMomentMatching(offset_amplitude_mean,
                                      offset_amplitude_stddev,
Philipp Haim's avatar
Philipp Haim committed
363
                                      prefix + 'zeromode',
Philipp Frank's avatar
Philipp Frank committed
364
                                      N)
Philipp Frank's avatar
fixup    
Philipp Frank committed
365
        if total_N > 0:
Martin Reinecke's avatar
Martin Reinecke committed
366
            zm = _Distributor(dofdex, zm.target, UnstructuredDomain(total_N)) @ zm
Philipp Haim's avatar
Philipp Haim committed
367
        return CorrelatedFieldMaker(zm, prefix, total_N)
368
369

    def add_fluctuations(self,
370
                         position_space,
371
372
373
374
375
376
377
378
                         fluctuations_mean,
                         fluctuations_stddev,
                         flexibility_mean,
                         flexibility_stddev,
                         asperity_mean,
                         asperity_stddev,
                         loglogavgslope_mean,
                         loglogavgslope_stddev,
Martin Reinecke's avatar
Martin Reinecke committed
379
380
381
382
                         prefix='',
                         index=None,
                         dofdex=None,
                         harmonic_partner=None):
Philipp Frank's avatar
Philipp Frank committed
383
384
        if harmonic_partner is None:
            harmonic_partner = position_space.get_default_codomain()
Philipp Frank's avatar
Fixup    
Philipp Frank committed
385
386
387
        else:
            position_space.check_codomain(harmonic_partner)
            harmonic_partner.check_codomain(position_space)
388

Philipp Haim's avatar
Philipp Haim committed
389
390
        if dofdex is None:
            dofdex = np.full(self._total_N, 0)
391
392
        elif len(dofdex) != self._total_N:
            raise ValueError("length of dofdex needs to match total_N")
Philipp Haim's avatar
Philipp Haim committed
393

Philipp Haim's avatar
Philipp Haim committed
394
        if self._total_N > 0:
Philipp Haim's avatar
Philipp Haim committed
395
            space = 1
Philipp Haim's avatar
Philipp Haim committed
396
397
            N = max(dofdex) + 1
            position_space = makeDomain((UnstructuredDomain(N), position_space))
Philipp Haim's avatar
Philipp Haim committed
398
399
        else:
            space = 0
Philipp Haim's avatar
Philipp Haim committed
400
            N = 0
Philipp Haim's avatar
Philipp Haim committed
401
            position_space = makeDomain(position_space)
Philipp Arras's avatar
Philipp Arras committed
402
        prefix = str(prefix)
Martin Reinecke's avatar
Martin Reinecke committed
403
        # assert isinstance(position_space[space], (RGSpace, HPSpace, GLSpace)
Philipp Arras's avatar
Philipp Arras committed
404

Philipp Arras's avatar
Philipp Arras committed
405
406
        fluct = _LognormalMomentMatching(fluctuations_mean,
                                         fluctuations_stddev,
407
                                         self._prefix + prefix + 'fluctuations',
Philipp Haim's avatar
Philipp Haim committed
408
                                         N)
Philipp Arras's avatar
Philipp Arras committed
409
        flex = _LognormalMomentMatching(flexibility_mean, flexibility_stddev,
410
                                        self._prefix + prefix + 'flexibility',
Philipp Haim's avatar
Philipp Haim committed
411
                                        N)
Philipp Arras's avatar
Philipp Arras committed
412
        asp = _LognormalMomentMatching(asperity_mean, asperity_stddev,
Martin Reinecke's avatar
Martin Reinecke committed
413
                                       self._prefix + prefix + 'asperity',
Philipp Haim's avatar
Philipp Haim committed
414
                                       N)
415
        avgsl = _normal(loglogavgslope_mean, loglogavgslope_stddev,
416
                        self._prefix + prefix + 'loglogavgslope', N)
Philipp Frank's avatar
Philipp Frank committed
417
        amp = _Amplitude(PowerSpace(harmonic_partner),
Martin Reinecke's avatar
Martin Reinecke committed
418
                         fluct, flex, asp, avgsl, self._azm,
Philipp Frank's avatar
fixup    
Philipp Frank committed
419
                         position_space[-1].total_volume,
420
                         self._prefix + prefix + 'spectrum', dofdex)
Philipp Haim's avatar
Philipp Haim committed
421

422
423
        if index is not None:
            self._a.insert(index, amp)
424
            self._position_spaces.insert(index, position_space)
425
426
        else:
            self._a.append(amp)
427
            self._position_spaces.append(position_space)
428

Philipp Frank's avatar
fixup    
Philipp Frank committed
429
    def _finalize_from_op(self):
Philipp Haim's avatar
Philipp Haim committed
430
        n_amplitudes = len(self._a)
Philipp Haim's avatar
Philipp Haim committed
431
        if self._total_N > 0:
Philipp Arras's avatar
Philipp Arras committed
432
433
434
            hspace = makeDomain(
                [UnstructuredDomain(self._total_N)] +
                [dd.target[-1].harmonic_partner for dd in self._a])
Philipp Haim's avatar
Philipp Haim committed
435
436
            spaces = tuple(range(1, n_amplitudes + 1))
            amp_space = 1
Philipp Haim's avatar
Philipp Haim committed
437
438
        else:
            hspace = makeDomain(
Philipp Arras's avatar
Philipp Arras committed
439
                [dd.target[0].harmonic_partner for dd in self._a])
Philipp Haim's avatar
Philipp Haim committed
440
            spaces = tuple(range(n_amplitudes))
Philipp Haim's avatar
Philipp Haim committed
441
            amp_space = 0
442

Martin Reinecke's avatar
Martin Reinecke committed
443
        expander = ContractionOperator(hspace, spaces=spaces).adjoint
Philipp Frank's avatar
fixup    
Philipp Frank committed
444
        azm = expander @ self._azm
445

446
        ht = HarmonicTransformOperator(hspace,
Philipp Haim's avatar
Philipp Haim committed
447
                                       self._position_spaces[0][amp_space],
Martin Reinecke's avatar
Martin Reinecke committed
448
                                       space=spaces[0])
449
        for i in range(1, n_amplitudes):
450
            ht = (HarmonicTransformOperator(ht.target,
Philipp Haim's avatar
Philipp Haim committed
451
                                            self._position_spaces[i][amp_space],
Martin Reinecke's avatar
Martin Reinecke committed
452
                                            space=spaces[i]) @ ht)
453

Philipp Haim's avatar
Philipp Haim committed
454
        pd = PowerDistributor(hspace, self._a[0].target[amp_space], amp_space)
455
        for i in range(1, n_amplitudes):
Philipp Arras's avatar
Philipp Arras committed
456
457
            pd = (pd @ PowerDistributor(
                pd.domain, self._a[i].target[amp_space], space=spaces[i]))
Philipp Arras's avatar
Philipp Arras committed
458

459
460
        a = ContractionOperator(pd.domain, spaces[1:]).adjoint @ self._a[0]
        for i in range(1, n_amplitudes):
Philipp Arras's avatar
Philipp Arras committed
461
            co = ContractionOperator(pd.domain, spaces[:i] + spaces[i + 1:])
462
            a = a*(co.adjoint @ self._a[i])
Philipp Arras's avatar
Philipp Arras committed
463

Philipp Frank's avatar
fixup    
Philipp Frank committed
464
        return ht(azm*(pd @ a)*ducktape(hspace, None, self._prefix + 'xi'))
Philipp Arras's avatar
Philipp Arras committed
465

Philipp Arras's avatar
Formats    
Philipp Arras committed
466
    def finalize(self, offset=None, prior_info=100):
Philipp Arras's avatar
Philipp Arras committed
467
468
469
        """
        offset vs zeromode: volume factor
        """
470
        op = self._finalize_from_op()
Philipp Arras's avatar
Philipp Arras committed
471
        if offset is not None:
472
473
474
475
476
477
478
            # Deviations from this offset must not be considered here as they
            # are learned by the zeromode
            if isinstance(offset, (Field, MultiField)):
                op = Adder(offset) @ op
            else:
                offset = float(offset)
                op = Adder(full(op.target, offset)) @ op
479
        self.statistics_summary(prior_info)
480
481
        return op

482
483
484
485
486
487
    def statistics_summary(self, prior_info):
        from ..sugar import from_random

        if prior_info == 0:
            return

488
489
        lst = [('Offset amplitude', self.amplitude_total_offset),
               ('Total fluctuation amplitude', self.total_fluctuation)]
490
        namps = len(self._a)
491
492
493
494
495
496
497
498
        if namps > 1:
            for ii in range(namps):
                lst.append(('Slice fluctuation (space {})'.format(ii),
                            self.slice_fluctuation(ii)))
                lst.append(('Average fluctuation (space {})'.format(ii),
                            self.average_fluctuation(ii)))

        for kk, op in lst:
499
500
501
502
503
            sc = StatCalculator()
            for _ in range(prior_info):
                sc.add(op(from_random('normal', op.domain)))
            mean = sc.mean.to_global_data()
            stddev = sc.var.sqrt().to_global_data()
504
            for m, s in zip(mean.flatten(), stddev.flatten()):
505
                logger.info('{}: {:.02E} ± {:.02E}'.format(kk, m, s))
506
507
508

    def moment_slice_to_average(self, fluctuations_slice_mean, nsamples=1000):
        fluctuations_slice_mean = float(fluctuations_slice_mean)
509
510
511
        if not fluctuations_slice_mean > 0:
            msg = "fluctuations_slice_mean must be greater zero; got {!r}"
            raise ValueError(msg.format(fluctuations_slice_mean))
512
        from ..sugar import from_random
513
514
        scm = 1.
        for a in self._a:
Philipp Haim's avatar
Philipp Haim committed
515
            op = a.fluctuation_amplitude*self._azm.one_over()
Martin Reinecke's avatar
Martin Reinecke committed
516
            res = np.array([op(from_random('normal', op.domain)).to_global_data()
517
518
                            for _ in range(nsamples)])
            scm *= res**2 + 1.
519
        return fluctuations_slice_mean/np.mean(np.sqrt(scm))
520

Philipp Arras's avatar
Philipp Arras committed
521
    @property
Philipp Haim's avatar
Philipp Haim committed
522
    def normalized_amplitudes(self):
523
        return self._a
Philipp Arras's avatar
Philipp Arras committed
524

Philipp Haim's avatar
Philipp Haim committed
525
526
527
528
529
530
531
    @property
    def amplitude(self):
        if len(self._a) > 1:
            s = ('If more than one spectrum is present in the model,',
                 ' no unique set of amplitudes exist because only the',
                 ' relative scale is determined.')
            raise NotImplementedError(s)
Philipp Haim's avatar
Fix    
Philipp Haim committed
532
533
        dom = self._a[0].target
        expand = ContractionOperator(dom, len(dom)-1).adjoint
Philipp Haim's avatar
Philipp Haim committed
534
535
        return self._a[0]*(expand @ self.amplitude_total_offset)

536
537
538
    @property
    def amplitude_total_offset(self):
        return self._azm
Philipp Arras's avatar
Philipp Arras committed
539
540

    @property
541
    def total_fluctuation(self):
542
        """Returns operator which acts on prior or posterior samples"""
543
        if len(self._a) == 0:
Philipp Arras's avatar
Philipp Arras committed
544
            raise NotImplementedError
545
        if len(self._a) == 1:
546
            return self.average_fluctuation(0)
547
548
        q = 1.
        for a in self._a:
Philipp Haim's avatar
Philipp Haim committed
549
            fl = a.fluctuation_amplitude*self._azm.one_over()
Philipp Arras's avatar
Philipp Arras committed
550
            q = q*(Adder(full(fl.target, 1.)) @ fl**2)
Philipp Arras's avatar
Formats    
Philipp Arras committed
551
        return (Adder(full(q.target, -1.)) @ q).sqrt()*self._azm
552

Philipp Arras's avatar
Philipp Arras committed
553
    def slice_fluctuation(self, space):
554
        """Returns operator which acts on prior or posterior samples"""
555
        if len(self._a) == 0:
Philipp Arras's avatar
Philipp Arras committed
556
            raise NotImplementedError
557
558
        if space >= len(self._a):
            raise ValueError(f"invalid space specified; got {space!r}")
559
        if len(self._a) == 1:
560
            return self.average_fluctuation(0)
561
562
        q = 1.
        for j in range(len(self._a)):
Philipp Haim's avatar
Philipp Haim committed
563
            fl = self._a[j].fluctuation_amplitude*self._azm.one_over()
564
            if j == space:
Philipp Arras's avatar
Philipp Arras committed
565
                q = q*fl**2
566
            else:
Philipp Arras's avatar
Philipp Arras committed
567
                q = q*(Adder(full(fl.target, 1.)) @ fl**2)
Philipp Arras's avatar
Formats    
Philipp Arras committed
568
        return q.sqrt()*self._azm
Philipp Arras's avatar
Philipp Arras committed
569
570

    def average_fluctuation(self, space):
571
        """Returns operator which acts on prior or posterior samples"""
572
        if len(self._a) == 0:
Philipp Arras's avatar
Philipp Arras committed
573
            raise NotImplementedError
574
575
        if space >= len(self._a):
            raise ValueError(f"invalid space specified; got {space!r}")
576
        if len(self._a) == 1:
Philipp Haim's avatar
Philipp Haim committed
577
578
            return self._a[0].fluctuation_amplitude
        return self._a[space].fluctuation_amplitude
579

580
581
    @staticmethod
    def offset_amplitude_realized(samples):
582
583
        res = 0.
        for s in samples:
Philipp Frank's avatar
fixes    
Philipp Frank committed
584
            res = res + s.mean()**2
585
        return np.sqrt(res/len(samples))
Philipp Arras's avatar
Philipp Arras committed
586

587
588
589
590
591
592
593
594
    @staticmethod
    def total_fluctuation_realized(samples):
        return _total_fluctuation_realized(samples)

    @staticmethod
    def slice_fluctuation_realized(samples, space):
        """Computes slice fluctuations from collection of field (defined in signal
        space) realizations."""
595
        ldom = len(samples[0].domain)
596
597
        if space >= ldom:
            raise ValueError(f"invalid space specified; got {space!r}")
598
        if ldom == 1:
599
            return _total_fluctuation_realized(samples)
Philipp Arras's avatar
Philipp Arras committed
600
        res1, res2 = 0., 0.
601
        for s in samples:
Philipp Frank's avatar
fixes    
Philipp Frank committed
602
603
604
605
606
607
608
            res1 = res1 + s**2
            res2 = res2 + s.mean(space)**2
        res1 = res1/len(samples)
        res2 = res2/len(samples)
        res = res1.mean() - res2.mean()
        return np.sqrt(res)

Philipp Arras's avatar
Philipp Arras committed
609
    @staticmethod
610
611
612
613
    def average_fluctuation_realized(samples, space):
        """Computes average fluctuations from collection of field (defined in signal
        space) realizations."""
        ldom = len(samples[0].domain)
614
615
        if space >= ldom:
            raise ValueError(f"invalid space specified; got {space!r}")
616
617
618
619
620
621
        if ldom == 1:
            return _total_fluctuation_realized(samples)
        spaces = ()
        for i in range(ldom):
            if i != space:
                spaces += (i,)
Philipp Arras's avatar
Philipp Arras committed
622
623
        res = 0.
        for s in samples:
624
625
626
627
            r = s.mean(spaces)
            res = res + (r - r.mean())**2
        res = res/len(samples)
        return np.sqrt(res.mean())