Skip to content
This repository
Fetching contributors…

Octocat-spinner-32-eaf2f5

Cannot retrieve contributors at this time

executable file 403 lines (350 sloc) 13.578 kb
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402
"""
Functions
---------
.. autosummary::
:toctree: generated/

fmin_l_bfgs_b

"""

## License for the Python wrapper
## ==============================

## Copyright (c) 2004 David M. Cooke <cookedm@physics.mcmaster.ca>

## Permission is hereby granted, free of charge, to any person obtaining a
## copy of this software and associated documentation files (the "Software"),
## to deal in the Software without restriction, including without limitation
## the rights to use, copy, modify, merge, publish, distribute, sublicense,
## and/or sell copies of the Software, and to permit persons to whom the
## Software is furnished to do so, subject to the following conditions:

## The above copyright notice and this permission notice shall be included in
## all copies or substantial portions of the Software.

## THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
## IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
## FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
## AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
## LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
## FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
## DEALINGS IN THE SOFTWARE.

## Modifications by Travis Oliphant and Enthought, Inc. for inclusion in SciPy

from __future__ import division, print_function, absolute_import

import numpy as np
from numpy import array, asarray, float64, int32, zeros
from . import _lbfgsb
from .optimize import approx_fprime, MemoizeJac, Result, _check_unknown_options

__all__ = ['fmin_l_bfgs_b']


def fmin_l_bfgs_b(func, x0, fprime=None, args=(),
                  approx_grad=0,
                  bounds=None, m=10, factr=1e7, pgtol=1e-5,
                  epsilon=1e-8,
                  iprint=-1, maxfun=15000, maxiter=15000, disp=None,
                  callback=None):
    """
Minimize a function func using the L-BFGS-B algorithm.

Parameters
----------
func : callable f(x,*args)
Function to minimise.
x0 : ndarray
Initial guess.
fprime : callable fprime(x,*args)
The gradient of `func`. If None, then `func` returns the function
value and the gradient (``f, g = func(x, *args)``), unless
`approx_grad` is True in which case `func` returns only ``f``.
args : sequence
Arguments to pass to `func` and `fprime`.
approx_grad : bool
Whether to approximate the gradient numerically (in which case
`func` returns only the function value).
bounds : list
``(min, max)`` pairs for each element in ``x``, defining
the bounds on that parameter. Use None for one of ``min`` or
``max`` when there is no bound in that direction.
m : int
The maximum number of variable metric corrections
used to define the limited memory matrix. (The limited memory BFGS
method does not store the full hessian but uses this many terms in an
approximation to it.)
factr : float
The iteration stops when
``(f^k - f^{k+1})/max{|f^k|,|f^{k+1}|,1} <= factr * eps``,
where ``eps`` is the machine precision, which is automatically
generated by the code. Typical values for `factr` are: 1e12 for
low accuracy; 1e7 for moderate accuracy; 10.0 for extremely
high accuracy.
pgtol : float
The iteration will stop when
``max{|proj g_i | i = 1, ..., n} <= pgtol``
where ``pg_i`` is the i-th component of the projected gradient.
epsilon : float
Step size used when `approx_grad` is True, for numerically
calculating the gradient
iprint : int
Controls the frequency of output. ``iprint < 0`` means no output;
``iprint == 0`` means write messages to stdout; ``iprint > 1`` in
addition means write logging information to a file named
``iterate.dat`` in the current working directory.
disp : int, optional
If zero, then no output. If a positive number, then this over-rides
`iprint` (i.e., `iprint` gets the value of `disp`).
maxfun : int
Maximum number of function evaluations.
maxiter : int
Maximum number of iterations.
callback : callable, optional
Called after each iteration, as ``callback(xk)``, where ``xk`` is the
current parameter vector.

Returns
-------
x : array_like
Estimated position of the minimum.
f : float
Value of `func` at the minimum.
d : dict
Information dictionary.

* d['warnflag'] is

- 0 if converged,
- 1 if too many function evaluations or too many iterations,
- 2 if stopped for another reason, given in d['task']

* d['grad'] is the gradient at the minimum (should be 0 ish)
* d['funcalls'] is the number of function calls made.
* d['nit'] is the number of iterations.

See also
--------
minimize: Interface to minimization algorithms for multivariate
functions. See the 'L-BFGS-B' `method` in particular.

Notes
-----
License of L-BFGS-B (FORTRAN code):

The version included here (in fortran code) is 3.0
(released April 25, 2011). It was written by Ciyou Zhu, Richard Byrd,
and Jorge Nocedal <nocedal@ece.nwu.edu>. It carries the following
condition for use:

This software is freely available, but we expect that all publications
describing work using this software, or all commercial products using it,
quote at least one of the references given below. This software is released
under the BSD License.

References
----------
* R. H. Byrd, P. Lu and J. Nocedal. A Limited Memory Algorithm for Bound
Constrained Optimization, (1995), SIAM Journal on Scientific and
Statistical Computing, 16, 5, pp. 1190-1208.
* C. Zhu, R. H. Byrd and J. Nocedal. L-BFGS-B: Algorithm 778: L-BFGS-B,
FORTRAN routines for large scale bound constrained optimization (1997),
ACM Transactions on Mathematical Software, 23, 4, pp. 550 - 560.
* J.L. Morales and J. Nocedal. L-BFGS-B: Remark on Algorithm 778: L-BFGS-B,
FORTRAN routines for large scale bound constrained optimization (2011),
ACM Transactions on Mathematical Software, 38, 1.

"""
    # handle fprime/approx_grad
    if approx_grad:
        fun = func
        jac = None
    elif fprime is None:
        fun = MemoizeJac(func)
        jac = fun.derivative
    else:
        fun = func
        jac = fprime

    # build options
    if disp is None:
        disp = iprint
    opts = {'disp': disp,
            'iprint': iprint,
            'maxcor': m,
            'ftol': factr * np.finfo(float).eps,
            'gtol': pgtol,
            'eps': epsilon,
            'maxfun': maxfun,
            'maxiter': maxiter,
            'callback': callback}

    res = _minimize_lbfgsb(fun, x0, args=args, jac=jac, bounds=bounds,
                           **opts)
    d = {'grad': res['jac'],
         'task': res['message'],
         'funcalls': res['nfev'],
         'nit': res['nit'],
         'warnflag': res['status']}
    f = res['fun']
    x = res['x']

    return x, f, d


def _minimize_lbfgsb(fun, x0, args=(), jac=None, bounds=None,
                     disp=None, maxcor=10, ftol=2.2204460492503131e-09,
                     gtol=1e-5, eps=1e-8, maxfun=15000, maxiter=15000,
                     iprint=-1, callback=None, **unknown_options):
    """
Minimize a scalar function of one or more variables using the L-BFGS-B
algorithm.

Options for the L-BFGS-B algorithm are:
disp : bool
Set to True to print convergence messages.
maxcor : int
The maximum number of variable metric corrections used to
define the limited memory matrix. (The limited memory BFGS
method does not store the full hessian but uses this many terms
in an approximation to it.)
factr : float
The iteration stops when ``(f^k -
f^{k+1})/max{|f^k|,|f^{k+1}|,1} <= factr * eps``, where ``eps``
is the machine precision, which is automatically generated by
the code. Typical values for `factr` are: 1e12 for low
accuracy; 1e7 for moderate accuracy; 10.0 for extremely high
accuracy.
gtol : float
The iteration will stop when ``max{|proj g_i | i = 1, ..., n}
<= gtol`` where ``pg_i`` is the i-th component of the
projected gradient.
eps : float
Step size used for numerical approximation of the jacobian.
disp : int
Set to True to print convergence messages.
maxfun : int
Maximum number of function evaluations.
maxiter : int
Maximum number of iterations.

This function is called by the `minimize` function with
`method=L-BFGS-B`. It is not supposed to be called directly.
"""
    _check_unknown_options(unknown_options)
    m = maxcor
    epsilon = eps
    pgtol = gtol
    factr = ftol / np.finfo(float).eps

    x0 = asarray(x0).ravel()
    n, = x0.shape

    if bounds is None:
        bounds = [(None, None)] * n
    if len(bounds) != n:
        raise ValueError('length of x0 != length of bounds')

    if disp is not None:
        if disp == 0:
            iprint = -1
        else:
            iprint = disp

    if jac is None:
        def func_and_grad(x):
            f = fun(x, *args)
            g = approx_fprime(x, fun, epsilon, *args)
            return f, g
    else:
        def func_and_grad(x):
            f = fun(x, *args)
            g = jac(x, *args)
            return f, g

    nbd = zeros(n, int32)
    low_bnd = zeros(n, float64)
    upper_bnd = zeros(n, float64)
    bounds_map = {(None, None): 0,
                  (1, None): 1,
                  (1, 1): 2,
                  (None, 1): 3}
    for i in range(0, n):
        l, u = bounds[i]
        if l is not None:
            low_bnd[i] = l
            l = 1
        if u is not None:
            upper_bnd[i] = u
            u = 1
        nbd[i] = bounds_map[l, u]

    x = array(x0, float64)
    f = array(0.0, float64)
    g = zeros((n,), float64)
    wa = zeros(2*m*n + 5*n + 11*m*m + 8*m, float64)
    iwa = zeros(3*n, int32)
    task = zeros(1, 'S60')
    csave = zeros(1, 'S60')
    lsave = zeros(4, int32)
    isave = zeros(44, int32)
    dsave = zeros(29, float64)

    task[:] = 'START'

    n_function_evals = 0
    n_iterations = 0

    while 1:
# x, f, g, wa, iwa, task, csave, lsave, isave, dsave = \
        _lbfgsb.setulb(m, x, low_bnd, upper_bnd, nbd, f, g, factr,
                       pgtol, wa, iwa, task, iprint, csave, lsave,
                       isave, dsave)
        task_str = task.tostring()
        if task_str.startswith(b'FG'):
            if n_function_evals > maxfun:
                task[:] = ('STOP: TOTAL NO. of f AND g EVALUATIONS '
                           'EXCEEDS LIMIT')
            else:
                # minimization routine wants f and g at the current x
                n_function_evals += 1
                # Overwrite f and g:
                f, g = func_and_grad(x)
        elif task_str.startswith(b'NEW_X'):
            # new iteration
            if n_iterations > maxiter:
                task[:] = 'STOP: TOTAL NO. of ITERATIONS EXCEEDS LIMIT'
            else:
                n_iterations += 1
                if callback is not None:
                    callback(x)
        else:
            break

    task_str = task.tostring().strip(b'\x00').strip()
    if task_str.startswith(b'CONV'):
        warnflag = 0
    elif n_function_evals > maxfun:
        warnflag = 1
    elif n_iterations > maxiter:
        warnflag = 1
    else:
        warnflag = 2

    return Result(fun=f, jac=g, nfev=n_function_evals, nit=n_iterations,
                  status=warnflag, message=task_str, x=x,
                  success=(warnflag == 0))


if __name__ == '__main__':
    def func(x):
        f = 0.25 * (x[0] - 1) ** 2
        for i in range(1, x.shape[0]):
            f += (x[i] - x[i-1] ** 2) ** 2
        f *= 4
        return f

    def grad(x):
        g = zeros(x.shape, float64)
        t1 = x[1] - x[0] ** 2
        g[0] = 2 * (x[0] - 1) - 16 * x[0] * t1
        for i in range(1, g.shape[0] - 1):
            t2 = t1
            t1 = x[i + 1] - x[i] ** 2
            g[i] = 8 * t2 - 16*x[i] * t1
        g[-1] = 8 * t1
        return g

    def func_and_grad(x):
        return func(x), grad(x)

    class Problem(object):
        def fun(self, x):
            return func_and_grad(x)

    factr = 1e7
    pgtol = 1e-5

    n = 25
    m = 10

    bounds = [(None, None)] * n
    for i in range(0, n, 2):
        bounds[i] = (1.0, 100)
    for i in range(1, n, 2):
        bounds[i] = (-100, 100)

    x0 = zeros((n,), float64)
    x0[:] = 3

    x, f, d = fmin_l_bfgs_b(func, x0, fprime=grad, m=m,
                            factr=factr, pgtol=pgtol)
    print(x)
    print(f)
    print(d)
    x, f, d = fmin_l_bfgs_b(func, x0, approx_grad=1,
                            m=m, factr=factr, pgtol=pgtol)
    print(x)
    print(f)
    print(d)
    x, f, d = fmin_l_bfgs_b(func_and_grad, x0, approx_grad=0,
                            m=m, factr=factr, pgtol=pgtol)
    print(x)
    print(f)
    print(d)
    p = Problem()
    x, f, d = fmin_l_bfgs_b(p.fun, x0, approx_grad=0,
                            m=m, factr=factr, pgtol=pgtol)
    print(x)
    print(f)
    print(d)
Something went wrong with that request. Please try again.