sugar.py 18.5 KB
Newer Older
1
2
3
4
5
6
7
8
9
10
11
12
# 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/>.
Theo Steininger's avatar
Theo Steininger committed
13
#
14
# Copyright(C) 2013-2020 Max-Planck-Society
Theo Steininger's avatar
Theo Steininger committed
15
#
16
# NIFTy is being developed at the Max-Planck-Institut fuer Astrophysik.
17

18
import sys
19
from time import time
20

21
import numpy as np
22

23
from . import pointwise, utilities
24
from .domain_tuple import DomainTuple
Martin Reinecke's avatar
Martin Reinecke committed
25
from .domains.power_space import PowerSpace
Martin Reinecke's avatar
cleanup    
Martin Reinecke committed
26
from .field import Field
27
from .logger import logger
Martin Reinecke's avatar
Martin Reinecke committed
28
29
from .multi_domain import MultiDomain
from .multi_field import MultiField
Philipp Arras's avatar
Philipp Arras committed
30
from .operators.block_diagonal_operator import BlockDiagonalOperator
Martin Reinecke's avatar
Martin Reinecke committed
31
from .operators.diagonal_operator import DiagonalOperator
Martin Reinecke's avatar
cleanup    
Martin Reinecke committed
32
from .operators.distributors import PowerDistributor
33
from .operators.operator import Operator
34
from .operators.sampling_enabler import SamplingDtypeSetter
35
from .operators.scaling_operator import ScalingOperator
Lukas Platz's avatar
Lukas Platz committed
36
from .plot import Plot
37

Martin Reinecke's avatar
step 1    
Martin Reinecke committed
38
39
__all__ = ['PS_field', 'power_analyze', 'create_power_operator',
           'create_harmonic_smoothing_operator', 'from_random',
Martin Reinecke's avatar
Martin Reinecke committed
40
           'full', 'makeField',
Martin Reinecke's avatar
Martin Reinecke committed
41
           'makeDomain', 'get_signal_variance', 'makeOp', 'domain_union',
Philipp Arras's avatar
Philipp Arras committed
42
           'get_default_codomain', 'single_plot', 'exec_time',
43
           'calculate_position'] + list(pointwise.ptw_dict.keys())
44

45

46
def PS_field(pspace, func):
Martin Reinecke's avatar
Martin Reinecke committed
47
48
49
50
51
52
53
54
    """Convenience function sampling a power spectrum

    Parameters
    ----------
    pspace : PowerSpace
        space at whose `k_lengths` the power spectrum function is evaluated
    func : function taking and returning a numpy.ndarray(float)
        the power spectrum function
Martin Reinecke's avatar
Martin Reinecke committed
55

Martin Reinecke's avatar
Martin Reinecke committed
56
57
    Returns
    -------
Martin Reinecke's avatar
Martin Reinecke committed
58
59
    Field
        A field defined on (pspace,) containing the computed function values
Martin Reinecke's avatar
Martin Reinecke committed
60
    """
Martin Reinecke's avatar
Martin Reinecke committed
61
62
    if not isinstance(pspace, PowerSpace):
        raise TypeError
Martin Reinecke's avatar
stage 1    
Martin Reinecke committed
63
    data = func(pspace.k_lengths)
64
    return Field(DomainTuple.make(pspace), data)
Martin Reinecke's avatar
Martin Reinecke committed
65

Martin Reinecke's avatar
Martin Reinecke committed
66

67
68
69
70
def get_signal_variance(spec, space):
    """
    Computes how much a field with a given power spectrum will vary in space

71
    This is a small helper function that computes the expected variance
72
73
74
75
76
77
78
79
    of a harmonically transformed sample of this power spectrum.

    Parameters
    ---------
    spec: method
        a method that takes one k-value and returns the power spectrum at that
        location
    space: PowerSpace or any harmonic Domain
Martin Reinecke's avatar
Martin Reinecke committed
80
81
82
83
        If this function is given a harmonic domain, it creates the naturally
        binned PowerSpace to that domain.
        The field, for which the signal variance is then computed, is assumed
        to have this PowerSpace as naturally binned PowerSpace
84
85
86
87
    """
    if space.harmonic:
        space = PowerSpace(space)
    if not isinstance(space, PowerSpace):
Martin Reinecke's avatar
Martin Reinecke committed
88
89
        raise ValueError(
            "space must be either a harmonic space or Power space.")
90
91
92
    field = PS_field(space, spec)
    dist = PowerDistributor(space.harmonic_partner, space)
    k_field = dist(field)
Martin Reinecke's avatar
Martin Reinecke committed
93
    return k_field.weight(2).s_sum()
94

95

96
97
def _single_power_analyze(field, idx, binbounds):
    power_domain = PowerSpace(field.domain[idx], binbounds)
Martin Reinecke's avatar
Martin Reinecke committed
98
99
    pd = PowerDistributor(field.domain, power_domain, idx)
    return pd.adjoint_times(field.weight(1)).weight(-1)  # divides by bin size
100
101


Martin Reinecke's avatar
Martin Reinecke committed
102
103
# MR FIXME: this function is not well suited for analyzing more than one
# subdomain at once, because it allows only one set of binbounds.
104
105
def power_analyze(field, spaces=None, binbounds=None,
                  keep_phase_information=False):
106
    """Computes the power spectrum for a subspace of `field`.
107
108

    Creates a PowerSpace for the space addressed by `spaces` with the given
109
    binning and computes the power spectrum as a :class:`Field` over this
110
    PowerSpace. This can only be done if the subspace to  be analyzed is a
Martin Reinecke's avatar
Martin Reinecke committed
111
112
    harmonic space. The resulting field has the same units as the square of the
    initial field.
113
114
115

    Parameters
    ----------
Philipp Arras's avatar
Fixups    
Philipp Arras committed
116
    field : Field
117
        The field to be analyzed
Martin Reinecke's avatar
Martin Reinecke committed
118
119
120
    spaces : None or int or tuple of int, optional
        The indices of subdomains for which the power spectrum shall be
        computed.
Martin Reinecke's avatar
Martin Reinecke committed
121
        If None, all subdomains will be converted.
122
        (default : None).
Martin Reinecke's avatar
Martin Reinecke committed
123
    binbounds : None or array-like, optional
124
        Inner bounds of the bins (default : None).
Martin Reinecke's avatar
Martin Reinecke committed
125
126
        if binbounds is None : bins are inferred.
    keep_phase_information : bool, optional
127
        If False, return a real-valued result containing the power spectrum
128
        of `field`.
129
        If True, return a complex-valued result whose real component
130
131
132
        contains the power spectrum computed from the real part of `field`,
        and whose imaginary component contains the power
        spectrum computed from the imaginary part of `field`.
133
134
135
136
137
138
        The absolute value of this result should be identical to the output
        of power_analyze with keep_phase_information=False.
        (default : False).

    Returns
    -------
Philipp Arras's avatar
Fixups    
Philipp Arras committed
139
    Field
140
        The output object. Its domain is a PowerSpace and it contains
Martin Reinecke's avatar
Martin Reinecke committed
141
        the power spectrum of `field`.
142
143
144
145
    """

    for sp in field.domain:
        if not sp.harmonic and not isinstance(sp, PowerSpace):
Martin Reinecke's avatar
Martin Reinecke committed
146
147
            logger.warning("WARNING: Field has a space in `domain` which is "
                           "neither harmonic nor a PowerSpace.")
148

149
    spaces = utilities.parse_spaces(spaces, len(field.domain))
150
151
152
153

    if len(spaces) == 0:
        raise ValueError("No space for analysis specified.")

Martin Reinecke's avatar
Martin Reinecke committed
154
    field_real = not utilities.iscomplextype(field.dtype)
155
156
157
    if (not field_real) and keep_phase_information:
        raise ValueError("cannot keep phase from real-valued input Field")

158
159
160
    if keep_phase_information:
        parts = [field.real*field.real, field.imag*field.imag]
    else:
161
162
163
164
        if field_real:
            parts = [field**2]
        else:
            parts = [field.real*field.real + field.imag*field.imag]
165
166

    for space_index in spaces:
Martin Reinecke's avatar
Martin Reinecke committed
167
        parts = [_single_power_analyze(part, space_index, binbounds)
168
169
170
171
172
                 for part in parts]

    return parts[0] + 1j*parts[1] if keep_phase_information else parts[0]


Martin Reinecke's avatar
cleanup    
Martin Reinecke committed
173
def _create_power_field(domain, power_spectrum):
Philipp Arras's avatar
Philipp Arras committed
174
    if not callable(power_spectrum):  # we have a Field defined on a PowerSpace
Martin Reinecke's avatar
tweaks    
Martin Reinecke committed
175
176
177
178
179
180
181
        if not isinstance(power_spectrum, Field):
            raise TypeError("Field object expected")
        if len(power_spectrum.domain) != 1:
            raise ValueError("exactly one domain required")
        if not isinstance(power_spectrum.domain[0], PowerSpace):
            raise TypeError("PowerSpace required")
        power_domain = power_spectrum.domain[0]
182
        fp = power_spectrum
Martin Reinecke's avatar
tweaks    
Martin Reinecke committed
183
184
    else:
        power_domain = PowerSpace(domain)
185
        fp = PS_field(power_domain, power_spectrum)
186

Martin Reinecke's avatar
Martin Reinecke committed
187
    return PowerDistributor(domain, power_domain)(fp)
188

189

190
def create_power_operator(domain, power_spectrum, space=None):
191
    """Creates a diagonal operator with the given power spectrum.
192

Philipp Arras's avatar
Philipp Arras committed
193
    Constructs a diagonal operator that is defined on the specified domain.
194

195
196
    Parameters
    ----------
Martin Reinecke's avatar
Martin Reinecke committed
197
    domain : Domain, tuple of Domain or DomainTuple
Philipp Arras's avatar
Philipp Arras committed
198
        Domain on which the power operator shall be defined.
Martin Reinecke's avatar
Martin Reinecke committed
199
200
    power_spectrum : callable or Field
        An object that contains the power spectrum as a function of k.
Martin Reinecke's avatar
Martin Reinecke committed
201
    space : int
Martin Reinecke's avatar
Martin Reinecke committed
202
        the domain index on which the power operator will work
Theo Steininger's avatar
Theo Steininger committed
203

204
205
    Returns
    -------
Martin Reinecke's avatar
Martin Reinecke committed
206
207
    DiagonalOperator
        An operator that implements the given power spectrum.
208
    """
Martin Reinecke's avatar
Martin Reinecke committed
209
    domain = DomainTuple.make(domain)
Martin Reinecke's avatar
Martin Reinecke committed
210
    space = utilities.infer_space(domain, space)
Martin Reinecke's avatar
Martin Reinecke committed
211
212
    field = _create_power_field(domain[space], power_spectrum)
    return DiagonalOperator(field, domain, space)
213

214

215
def create_harmonic_smoothing_operator(domain, space, sigma):
Martin Reinecke's avatar
Martin Reinecke committed
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
    """Creates an operator which smoothes a subspace of a harmonic domain.

    Parameters
    ----------
    domain: DomainTuple
        The total domain and target of the operator
    space : int
        the index of the subspace on which the operator acts.
        This must be a harmonic space
    sigma : float
        The sigma of the Gaussian smoothing kernel

    Returns
    -------
    DiagonalOperator
        The requested smoothing operator
    """
233
234
235
    kfunc = domain[space].get_fft_smoothing_kernel_function(sigma)
    return DiagonalOperator(kfunc(domain[space].get_k_length_array()), domain,
                            space)
Martin Reinecke's avatar
step 1    
Martin Reinecke committed
236
237
238


def full(domain, val):
Martin Reinecke's avatar
Martin Reinecke committed
239
240
241
242
243
244
245
246
247
248
249
    """Convenience function creating Fields/MultiFields with uniform values.

    Parameters
    ----------
    domain : Domainoid
        the intended domain of the output field
    val : scalar value
        the uniform value to be placed into all entries of the result

    Returns
    -------
Martin Reinecke's avatar
Martin Reinecke committed
250
251
    Field or MultiField
        The newly created uniform field
Martin Reinecke's avatar
Martin Reinecke committed
252
    """
Martin Reinecke's avatar
step 1    
Martin Reinecke committed
253
254
255
256
257
    if isinstance(domain, (dict, MultiDomain)):
        return MultiField.full(domain, val)
    return Field.full(domain, val)


258
def from_random(domain, random_type='normal', dtype=np.float64, **kwargs):
Martin Reinecke's avatar
Martin Reinecke committed
259
260
261
262
263
264
    """Convenience function creating Fields/MultiFields with random values.

    Parameters
    ----------
    domain : Domainoid
        the intended domain of the output field
265
266
    random_type : 'pm1', 'normal', or 'uniform'
            The random distribution to use.
Martin Reinecke's avatar
Martin Reinecke committed
267
268
    dtype : type
        data type of the output field (e.g. numpy.float64)
269
270
        If the datatype is complex, each real an imaginary part have
        variance 1.
Martin Reinecke's avatar
Martin Reinecke committed
271
272
273
274
275
    **kwargs : additional parameters for the random distribution
        ('mean' and 'std' for 'normal', 'low' and 'high' for 'uniform')

    Returns
    -------
Martin Reinecke's avatar
Martin Reinecke committed
276
277
    Field or MultiField
        The newly created random field
278
279
280

    Notes
    -----
Martin Reinecke's avatar
Martin Reinecke committed
281
282
283
284
    When called with a multi-domain, the individual fields will be drawn in
    alphabetical order of the multi-domain's domain keys. As a consequence,
    renaming these keys may cause the multi-field to be filled with different
    random numbers, even for the same initial RNG state.
Martin Reinecke's avatar
Martin Reinecke committed
285
    """
Martin Reinecke's avatar
step 1    
Martin Reinecke committed
286
    if isinstance(domain, (dict, MultiDomain)):
287
288
        return MultiField.from_random(domain, random_type, dtype, **kwargs)
    return Field.from_random(domain, random_type, dtype, **kwargs)
Martin Reinecke's avatar
step 1    
Martin Reinecke committed
289
290


Martin Reinecke's avatar
Martin Reinecke committed
291
def makeField(domain, arr):
Martin Reinecke's avatar
Martin Reinecke committed
292
293
294
295
296
297
298
299
300
301
302
303
    """Convenience function creating Fields/MultiFields from Numpy arrays or
    dicts of Numpy arrays.

    Parameters
    ----------
    domain : Domainoid
        the intended domain of the output field
    arr : Numpy array if `domain` corresponds to a `DomainTuple`,
          dictionary of Numpy arrays if `domain` corresponds to a `MultiDomain`

    Returns
    -------
Martin Reinecke's avatar
Martin Reinecke committed
304
305
    Field or MultiField
        The newly created random field
Martin Reinecke's avatar
Martin Reinecke committed
306
    """
Martin Reinecke's avatar
step 1    
Martin Reinecke committed
307
    if isinstance(domain, (dict, MultiDomain)):
Martin Reinecke's avatar
Martin Reinecke committed
308
309
        return MultiField.from_raw(domain, arr)
    return Field.from_raw(domain, arr)
Martin Reinecke's avatar
step 1    
Martin Reinecke committed
310
311
312


def makeDomain(domain):
Martin Reinecke's avatar
Martin Reinecke committed
313
314
315
316
    """Convenience function creating DomainTuples/MultiDomains Domainoids.

    Parameters
    ----------
Martin Reinecke's avatar
Martin Reinecke committed
317
    domain : Domainoid (can be DomainTuple, MultiDomain, dict, Domain or list of Domains)
Martin Reinecke's avatar
Martin Reinecke committed
318
319
320
321
        the description of the requested (multi-)domain

    Returns
    -------
Martin Reinecke's avatar
Martin Reinecke committed
322
323
    DomainTuple or MultiDomain
        The newly created domain object
Martin Reinecke's avatar
Martin Reinecke committed
324
    """
Martin Reinecke's avatar
fixes    
Martin Reinecke committed
325
    if isinstance(domain, (MultiDomain, dict)):
Martin Reinecke's avatar
step 1    
Martin Reinecke committed
326
327
        return MultiDomain.make(domain)
    return DomainTuple.make(domain)
328
329


330
def makeOp(input, dom=None):
Martin Reinecke's avatar
Martin Reinecke committed
331
332
333
334
335
336
    """Converts a Field or MultiField to a diagonal operator.

    Parameters
    ----------
    input : None, Field or MultiField
        - if None, None is returned.
337
338
        - if Field on scalar-domain, a ScalingOperator with the coefficient
            given by the Field is returned.
Martin Reinecke's avatar
Martin Reinecke committed
339
340
341
342
343
        - if Field, a DiagonalOperator with the coefficients given by this
            Field is returned.
        - if MultiField, a BlockDiagonalOperator with entries given by this
            MultiField is returned.

344
345
346
    dom : DomainTuple or MultiDomain
        if `input` is a scalar, this is used as the operator's domain

Martin Reinecke's avatar
Martin Reinecke committed
347
348
349
350
    Notes
    -----
    No volume factors are applied.
    """
Martin Reinecke's avatar
Martin Reinecke committed
351
352
    if input is None:
        return None
353
    if np.isscalar(input):
Rouven Lemmerz's avatar
Typo    
Rouven Lemmerz committed
354
        if not isinstance(dom, (DomainTuple, MultiDomain)):
355
            raise TypeError("need proper `dom` argument")
356
        return ScalingOperator(dom, input)
357
358
359
    if dom is not None:
        if not dom == input.domain:
            raise ValueError("domain mismatch")
360
    if input.domain is DomainTuple.scalar_domain():
361
        return ScalingOperator(input.domain, input.val[()])
Martin Reinecke's avatar
Martin Reinecke committed
362
363
364
    if isinstance(input, Field):
        return DiagonalOperator(input)
    if isinstance(input, MultiField):
Martin Reinecke's avatar
Martin Reinecke committed
365
        return BlockDiagonalOperator(
Martin Reinecke's avatar
fix    
Martin Reinecke committed
366
            input.domain, {key: makeOp(val) for key, val in input.items()})
Martin Reinecke's avatar
Martin Reinecke committed
367
368
    raise NotImplementedError

Martin Reinecke's avatar
more    
Martin Reinecke committed
369
370

def domain_union(domains):
Martin Reinecke's avatar
Martin Reinecke committed
371
372
373
374
375
376
377
378
    """Computes the union of multiple DomainTuples/MultiDomains.

    Parameters
    ----------
    domains : list of DomainTuple or MultiDomain
        - if DomainTuple, all entries must be equal
        - if MultiDomain, there must not be any conflicting components
    """
Martin Reinecke's avatar
more    
Martin Reinecke committed
379
    if isinstance(domains[0], DomainTuple):
Martin Reinecke's avatar
Martin Reinecke committed
380
        if any(dom != domains[0] for dom in domains[1:]):
Martin Reinecke's avatar
more    
Martin Reinecke committed
381
382
383
384
            raise ValueError("domain mismatch")
        return domains[0]
    return MultiDomain.union(domains)

Martin Reinecke's avatar
more    
Martin Reinecke committed
385

386
387
388
389
390
391
392
393
394
395
# Pointwise functions

_current_module = sys.modules[__name__]

for f in pointwise.ptw_dict.keys():
    def func(f):
        def func2(x, *args, **kwargs):
           return x.ptw(f, *args, **kwargs)
        return func2
    setattr(_current_module, f, func(f))
Martin Reinecke's avatar
Martin Reinecke committed
396
397


398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
def get_default_codomain(domainoid, space=None):
    """For `RGSpace`, returns the harmonic partner domain.
    For `DomainTuple`, returns a copy of the object in which the domain
    indexed by `space` is substituted by its harmonic partner domain.
    In this case, if `space` is None, it is set to 0 if the `DomainTuple`
    contains exactly one domain.

    Parameters
    ----------
    domain: `RGSpace` or `DomainTuple`
        Domain for which to constuct the default harmonic partner
    space: int
        Optional index of the subdomain to be replaced by its default
        codomain. `domain[space]` must be of class `RGSpace`.
    """
    from .domains.rg_space import RGSpace
414
415
416
    from .domains.hp_space import HPSpace
    from .domains.gl_space import GLSpace
    from .domains.lm_space import LMSpace
417
418
419
420
421
422
    if isinstance(domainoid, RGSpace):
        return domainoid.get_default_codomain()
    if not isinstance(domainoid, DomainTuple):
        raise TypeError(
            'Works only on RGSpaces and DomainTuples containing those')
    space = utilities.infer_space(domainoid, space)
423
424
    if not isinstance(domainoid[space], (RGSpace, HPSpace, GLSpace, LMSpace)):
        raise TypeError("can only codomain structrued spaces")
425
426
427
    ret = [dom for dom in domainoid]
    ret[space] = domainoid[space].get_default_codomain()
    return DomainTuple.make(ret)
Lukas Platz's avatar
Lukas Platz committed
428
429
430
431
432
433
434
435
436
437
438


def single_plot(field, **kwargs):
    """Creates a single plot using `Plot`.
    Keyword arguments are passed to both `Plot.add` and `Plot.output`.
    """
    p = Plot()
    p.add(field, **kwargs)
    if 'title' in kwargs:
        del(kwargs['title'])
    p.output(**kwargs)
439
440
441
442


def exec_time(obj, want_metric=True):
    """Times the execution time of an operator or an energy."""
Philipp Arras's avatar
Philipp Arras committed
443
444
445
    from .linearization import Linearization
    from .minimization.energy import Energy
    from .operators.energy_operators import EnergyOperator
446
447
448
    if isinstance(obj, Energy):
        t0 = time()
        obj.at(0.99*obj.position)
449
        logger.info('Energy.at(): {}'.format(time() - t0))
450
451
452

        t0 = time()
        obj.value
453
        logger.info('Energy.value: {}'.format(time() - t0))
454
455
        t0 = time()
        obj.gradient
456
        logger.info('Energy.gradient: {}'.format(time() - t0))
457
458
        t0 = time()
        obj.metric
459
        logger.info('Energy.metric: {}'.format(time() - t0))
460
461
462

        t0 = time()
        obj.apply_metric(obj.position)
463
        logger.info('Energy.apply_metric: {}'.format(time() - t0))
464
465
466

        t0 = time()
        obj.metric(obj.position)
467
        logger.info('Energy.metric(position): {}'.format(time() - t0))
468
469
    elif isinstance(obj, Operator):
        want_metric = bool(want_metric)
470
        pos = from_random(obj.domain, 'normal')
471
472
        t0 = time()
        obj(pos)
473
        logger.info('Operator call with field: {}'.format(time() - t0))
474
475
476
477

        lin = Linearization.make_var(pos, want_metric=want_metric)
        t0 = time()
        res = obj(lin)
478
        logger.info('Operator call with linearization: {}'.format(time() - t0))
479
480
481
482

        if isinstance(obj, EnergyOperator):
            t0 = time()
            res.gradient
483
            logger.info('Gradient evaluation: {}'.format(time() - t0))
484
485
486
487

            if want_metric:
                t0 = time()
                res.metric(pos)
488
                logger.info('Metric apply: {}'.format(time() - t0))
489
490
    else:
        raise TypeError
Philipp Arras's avatar
Philipp Arras committed
491
492
493
494


def calculate_position(operator, output):
    """Finds approximate preimage of an operator for a given output."""
Philipp Arras's avatar
Philipp Arras committed
495
496
497
498
499
    from .minimization.descent_minimizers import NewtonCG
    from .minimization.iteration_controllers import GradientNormController
    from .minimization.metric_gaussian_kl import MetricGaussianKL
    from .operators.scaling_operator import ScalingOperator
    from .operators.energy_operators import GaussianEnergy, StandardHamiltonian
Philipp Arras's avatar
Philipp Arras committed
500
501
502
503
    if not isinstance(operator, Operator):
        raise TypeError
    if output.domain != operator.target:
        raise TypeError
504
    if isinstance(output, MultiField):
505
        cov = 1e-3*max([np.max(np.abs(vv)) for vv in output.val.values()])**2
Philipp Arras's avatar
Philipp Arras committed
506
507
508
509
510
        invcov = ScalingOperator(output.domain, cov).inverse
        dtype = list(set([ff.dtype for ff in output.values()]))
        if len(dtype) != 1:
            raise ValueError('Only MultiFields with one dtype supported.')
        dtype = dtype[0]
511
    else:
512
        cov = 1e-3*np.max(np.abs(output.val))**2
Philipp Arras's avatar
Philipp Arras committed
513
        dtype = output.dtype
514
    invcov = ScalingOperator(output.domain, cov).inverse
515
516
517
    invcov = SamplingDtypeSetter(invcov, output.dtype)
    invcov = SamplingDtypeSetter(invcov, output.dtype)
    d = output + invcov.draw_sample(from_inverse=True)
Philipp Arras's avatar
Philipp Arras committed
518
    lh = GaussianEnergy(d, invcov) @ operator
Philipp Arras's avatar
Philipp Arras committed
519
520
    H = StandardHamiltonian(
        lh, ic_samp=GradientNormController(iteration_limit=200))
521
    pos = 0.1*from_random(operator.domain)
Philipp Arras's avatar
Philipp Arras committed
522
    minimizer = NewtonCG(GradientNormController(iteration_limit=10, name='findpos'))
Philipp Arras's avatar
Philipp Arras committed
523
    for ii in range(3):
Philipp Arras's avatar
Philipp Arras committed
524
        logger.info(f'Start iteration {ii+1}/3')
525
        kl = MetricGaussianKL.make(pos, H, 3, mirror_samples=True)
Philipp Arras's avatar
Philipp Arras committed
526
527
528
        kl, _ = minimizer(kl)
        pos = kl.position
    return pos