scipy_minimizer.py 6.24 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/>.
#
Martin Reinecke's avatar
Martin Reinecke committed
14
# Copyright(C) 2013-2018 Max-Planck-Society
Martin Reinecke's avatar
Martin Reinecke committed
15
16
17
18
#
# NIFTy is being developed at the Max-Planck-Institut fuer Astrophysik
# and financially supported by the Studienstiftung des deutschen Volkes.

19
from __future__ import absolute_import, division, print_function
20

21
import numpy as np
Philipp Arras's avatar
Philipp Arras committed
22

23
from .. import dobj
24
from ..compat import *
Martin Reinecke's avatar
Martin Reinecke committed
25
from ..field import Field
Martin Reinecke's avatar
Martin Reinecke committed
26
from ..logger import logger
Philipp Arras's avatar
Philipp Arras committed
27
28
from ..multi_field import MultiField
from ..utilities import iscomplextype
Martin Reinecke's avatar
fix    
Martin Reinecke committed
29
from .iteration_controllers import IterationController
30
from .minimizer import Minimizer
31
32


33
def _multiToArray(fld):
Martin Reinecke's avatar
Martin Reinecke committed
34
35
    szall = sum(2*v.size if iscomplextype(v.dtype) else v.size
                for v in fld.values())
36
37
38
    res = np.empty(szall, dtype=np.float64)
    ofs = 0
    for val in fld.values():
39
40
41
        sz2 = 2*val.size if iscomplextype(val.dtype) else val.size
        locdat = val.local_data.reshape(-1)
        if iscomplextype(val.dtype):
Martin Reinecke's avatar
Martin Reinecke committed
42
            locdat = locdat.view(locdat.real.dtype)
43
44
        res[ofs:ofs+sz2] = locdat
        ofs += sz2
45
46
47
    return res


Martin Reinecke's avatar
Martin Reinecke committed
48
def _toArray(fld):
49
50
51
    if isinstance(fld, Field):
        return fld.local_data.reshape(-1)
    return _multiToArray(fld)
Martin Reinecke's avatar
Martin Reinecke committed
52
53


Martin Reinecke's avatar
Martin Reinecke committed
54
def _toArray_rw(fld):
55
56
57
    if isinstance(fld, Field):
        return fld.local_data.copy().reshape(-1)
    return _multiToArray(fld)
Martin Reinecke's avatar
Martin Reinecke committed
58
59


60
61
62
def _toField(arr, template):
    if isinstance(template, Field):
        return Field.from_local_data(template.domain,
Martin Reinecke's avatar
fix    
Martin Reinecke committed
63
                                     arr.reshape(template.shape).copy())
64
65
    ofs = 0
    res = []
66
67
68
69
70
71
72
73
    for v in template.values():
        sz2 = 2*v.size if iscomplextype(v.dtype) else v.size
        locdat = arr[ofs:ofs+sz2].copy()
        if iscomplextype(v.dtype):
            locdat = locdat.view(np.complex128)
        res.append(Field.from_local_data(v.domain, locdat.reshape(v.shape)))
        ofs += sz2
    return MultiField(template.domain, tuple(res))
Martin Reinecke's avatar
Martin Reinecke committed
74
75


76
77
78
79
80
81
class _MinHelper(object):
    def __init__(self, energy):
        self._energy = energy
        self._domain = energy.position.domain

    def _update(self, x):
82
        pos = _toField(x, self._energy.position)
Martin Reinecke's avatar
Martin Reinecke committed
83
        if (pos != self._energy.position).any():
84
            self._energy = self._energy.at(pos)
85
86
87
88
89
90
91

    def fun(self, x):
        self._update(x)
        return self._energy.value

    def jac(self, x):
        self._update(x)
Martin Reinecke's avatar
Martin Reinecke committed
92
        return _toArray_rw(self._energy.gradient)
93
94
95

    def hessp(self, x, p):
        self._update(x)
96
        res = self._energy.metric(_toField(p, self._energy.position))
Martin Reinecke's avatar
Martin Reinecke committed
97
        return _toArray_rw(res)
Martin Reinecke's avatar
Martin Reinecke committed
98
99
100
101
102
103
104
105
106
107
108
109
110


class ScipyMinimizer(Minimizer):
    """Scipy-based minimizer

    Parameters
    ----------
    method     : str
        The selected Scipy minimization method.
    options    : dictionary
        A set of custom options for the selected minimizer.
    """

111
    def __init__(self, method, options, need_hessp, bounds):
Martin Reinecke's avatar
Martin Reinecke committed
112
113
114
115
116
        if not dobj.is_numpy():
            raise NotImplementedError
        self._method = method
        self._options = options
        self._need_hessp = need_hessp
117
        self._bounds = bounds
Martin Reinecke's avatar
Martin Reinecke committed
118
119
120

    def __call__(self, energy):
        import scipy.optimize as opt
121
122
123
124
125
126
127
128
        hlp = _MinHelper(energy)
        energy = None  # drop handle, since we don't need it any more
        bounds = None
        if self._bounds is not None:
            if len(self._bounds) == 2:
                lo = self._bounds[0]
                hi = self._bounds[1]
                bounds = [(lo, hi)]*hlp._energy.position.size
Martin Reinecke's avatar
Martin Reinecke committed
129
            else:
130
131
                raise ValueError("unrecognized bounds")

Martin Reinecke's avatar
Martin Reinecke committed
132
        x = _toArray_rw(hlp._energy.position)
133
134
135
        hessp = hlp.hessp if self._need_hessp else None
        r = opt.minimize(hlp.fun, x, method=self._method, jac=hlp.jac,
                         hessp=hessp, options=self._options, bounds=bounds)
Martin Reinecke's avatar
Martin Reinecke committed
136
        if not r.success:
137
            logger.error("Problem in Scipy minimization: {}".format(r.message))
138
139
            return hlp._energy, IterationController.ERROR
        return hlp._energy, IterationController.CONVERGED
Martin Reinecke's avatar
Martin Reinecke committed
140
141


142
def L_BFGS_B(ftol, gtol, maxiter, maxcor=10, disp=False, bounds=None):
Martin Reinecke's avatar
Martin Reinecke committed
143
144
145
146
147
148
    """Returns a ScipyMinimizer object carrying out the L-BFGS-B algorithm.

    See Also
    --------
    ScipyMinimizer
    """
149
    options = {"ftol": ftol, "gtol": gtol, "maxiter": maxiter,
Martin Reinecke's avatar
fix    
Martin Reinecke committed
150
               "maxcor": maxcor, "disp": disp}
151
    return ScipyMinimizer("L-BFGS-B", options, False, bounds)
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171


class ScipyCG(Minimizer):
    def __init__(self, tol, maxiter):
        if not dobj.is_numpy():
            raise NotImplementedError
        self._tol = tol
        self._maxiter = maxiter

    def __call__(self, energy, preconditioner=None):
        from scipy.sparse.linalg import LinearOperator as scipy_linop, cg
        from .quadratic_energy import QuadraticEnergy
        if not isinstance(energy, QuadraticEnergy):
            raise ValueError("need a quadratic energy for CG")

        class mymatvec(object):
            def __init__(self, op):
                self._op = op

            def __call__(self, inp):
172
                return _toArray(self._op(_toField(inp, energy.position)))
173
174

        op = energy._A
Martin Reinecke's avatar
Martin Reinecke committed
175
176
        b = _toArray(energy._b)
        sx = _toArray(energy.position)
177
178
179
180
181
182
183
184
185
186
        sci_op = scipy_linop(shape=(op.domain.size, op.target.size),
                             matvec=mymatvec(op))
        prec_op = None
        if preconditioner is not None:
            prec_op = scipy_linop(shape=(op.domain.size, op.target.size),
                                  matvec=mymatvec(preconditioner))
        res, stat = cg(sci_op, b, x0=sx, tol=self._tol, M=prec_op,
                       maxiter=self._maxiter)
        stat = (IterationController.CONVERGED if stat >= 0 else
                IterationController.ERROR)
187
        return energy.at(_toField(res, energy.position)), stat