Submit
Path:
~
/
/
proc
/
thread-self
/
root
/
opt
/
alt
/
python35
/
lib64
/
python3.5
/
site-packages
/
scipy
/
optimize
/
__pycache__
/
File Content:
_differentialevolution.cpython-35.pyc
��W. � @ s� d Z d d l m Z m Z m Z d d l Z d d l m Z m Z d d l m Z d d l Z d g Z e j e j � j Z f d d d d d d f d d d d d d d d � Z Gd d � d e � Z d d � Z d S)zn differential_evolution: The differential evolution global optimization algorithm Added by Andrew Nelson 2014 � )�division�print_function�absolute_importN)�OptimizeResult�minimize)�_status_message�differential_evolution�best1bini� � g{�G�z�?g �?� gffffff�?FT�latinhypercubec C sa t | | d | d | d | d | d | d | d | d | d | d | d | d | � } | j � S) a�! Finds the global minimum of a multivariate function. Differential Evolution is stochastic in nature (does not use gradient methods) to find the minimium, and can search large areas of candidate space, but often requires larger numbers of function evaluations than conventional gradient based techniques. The algorithm is due to Storn and Price [1]_. Parameters ---------- func : callable The objective function to be minimized. Must be in the form ``f(x, *args)``, where ``x`` is the argument in the form of a 1-D array and ``args`` is a tuple of any additional fixed parameters needed to completely specify the function. bounds : sequence Bounds for variables. ``(min, max)`` pairs for each element in ``x``, defining the lower and upper bounds for the optimizing argument of `func`. It is required to have ``len(bounds) == len(x)``. ``len(bounds)`` is used to determine the number of parameters in ``x``. args : tuple, optional Any additional fixed parameters needed to completely specify the objective function. strategy : str, optional The differential evolution strategy to use. Should be one of: - 'best1bin' - 'best1exp' - 'rand1exp' - 'randtobest1exp' - 'best2exp' - 'rand2exp' - 'randtobest1bin' - 'best2bin' - 'rand2bin' - 'rand1bin' The default is 'best1bin'. maxiter : int, optional The maximum number of generations over which the entire population is evolved. The maximum number of function evaluations (with no polishing) is: ``(maxiter + 1) * popsize * len(x)`` popsize : int, optional A multiplier for setting the total population size. The population has ``popsize * len(x)`` individuals. tol : float, optional When the mean of the population energies, multiplied by tol, divided by the standard deviation of the population energies is greater than 1 the solving process terminates: ``convergence = mean(pop) * tol / stdev(pop) > 1`` mutation : float or tuple(float, float), optional The mutation constant. In the literature this is also known as differential weight, being denoted by F. If specified as a float it should be in the range [0, 2]. If specified as a tuple ``(min, max)`` dithering is employed. Dithering randomly changes the mutation constant on a generation by generation basis. The mutation constant for that generation is taken from ``U[min, max)``. Dithering can help speed convergence significantly. Increasing the mutation constant increases the search radius, but will slow down convergence. recombination : float, optional The recombination constant, should be in the range [0, 1]. In the literature this is also known as the crossover probability, being denoted by CR. Increasing this value allows a larger number of mutants to progress into the next generation, but at the risk of population stability. seed : int or `np.random.RandomState`, optional If `seed` is not specified the `np.RandomState` singleton is used. If `seed` is an int, a new `np.random.RandomState` instance is used, seeded with seed. If `seed` is already a `np.random.RandomState instance`, then that `np.random.RandomState` instance is used. Specify `seed` for repeatable minimizations. disp : bool, optional Display status messages callback : callable, `callback(xk, convergence=val)`, optional A function to follow the progress of the minimization. ``xk`` is the current value of ``x0``. ``val`` represents the fractional value of the population convergence. When ``val`` is greater than one the function halts. If callback returns `True`, then the minimization is halted (any polishing is still carried out). polish : bool, optional If True (default), then `scipy.optimize.minimize` with the `L-BFGS-B` method is used to polish the best population member at the end, which can improve the minimization slightly. init : string, optional Specify how the population initialization is performed. Should be one of: - 'latinhypercube' - 'random' The default is 'latinhypercube'. Latin Hypercube sampling tries to maximize coverage of the available parameter space. 'random' initializes the population randomly - this has the drawback that clustering can occur, preventing the whole of parameter space being covered. Returns ------- res : OptimizeResult The optimization result represented as a `OptimizeResult` object. Important attributes are: ``x`` the solution array, ``success`` a Boolean flag indicating if the optimizer exited successfully and ``message`` which describes the cause of the termination. See `OptimizeResult` for a description of other attributes. If `polish` was employed, and a lower minimum was obtained by the polishing, then OptimizeResult also contains the ``jac`` attribute. Notes ----- Differential evolution is a stochastic population based method that is useful for global optimization problems. At each pass through the population the algorithm mutates each candidate solution by mixing with other candidate solutions to create a trial candidate. There are several strategies [2]_ for creating trial candidates, which suit some problems more than others. The 'best1bin' strategy is a good starting point for many systems. In this strategy two members of the population are randomly chosen. Their difference is used to mutate the best member (the `best` in `best1bin`), :math:`b_0`, so far: .. math:: b' = b_0 + mutation * (population[rand0] - population[rand1]) A trial vector is then constructed. Starting with a randomly chosen 'i'th parameter the trial is sequentially filled (in modulo) with parameters from `b'` or the original candidate. The choice of whether to use `b'` or the original candidate is made with a binomial distribution (the 'bin' in 'best1bin') - a random number in [0, 1) is generated. If this number is less than the `recombination` constant then the parameter is loaded from `b'`, otherwise it is loaded from the original candidate. The final parameter is always loaded from `b'`. Once the trial candidate is built its fitness is assessed. If the trial is better than the original candidate then it takes its place. If it is also better than the best overall candidate it also replaces that. To improve your chances of finding a global minimum use higher `popsize` values, with higher `mutation` and (dithering), but lower `recombination` values. This has the effect of widening the search radius, but slowing convergence. .. versionadded:: 0.15.0 Examples -------- Let us consider the problem of minimizing the Rosenbrock function. This function is implemented in `rosen` in `scipy.optimize`. >>> from scipy.optimize import rosen, differential_evolution >>> bounds = [(0,2), (0, 2), (0, 2), (0, 2), (0, 2)] >>> result = differential_evolution(rosen, bounds) >>> result.x, result.fun (array([1., 1., 1., 1., 1.]), 1.9216496320061384e-19) Next find the minimum of the Ackley function (http://en.wikipedia.org/wiki/Test_functions_for_optimization). >>> from scipy.optimize import differential_evolution >>> import numpy as np >>> def ackley(x): ... arg1 = -0.2 * np.sqrt(0.5 * (x[0] ** 2 + x[1] ** 2)) ... arg2 = 0.5 * (np.cos(2. * np.pi * x[0]) + np.cos(2. * np.pi * x[1])) ... return -20. * np.exp(arg1) - np.exp(arg2) + 20. + np.e >>> bounds = [(-5, 5), (-5, 5)] >>> result = differential_evolution(ackley, bounds) >>> result.x, result.fun (array([ 0., 0.]), 4.4408920985006262e-16) References ---------- .. [1] Storn, R and Price, K, Differential Evolution - a Simple and Efficient Heuristic for Global Optimization over Continuous Spaces, Journal of Global Optimization, 1997, 11, 341 - 359. .. [2] http://www1.icsi.berkeley.edu/~storn/code.html .. [3] http://en.wikipedia.org/wiki/Differential_evolution �args�strategy�maxiter�popsize�tol�mutation� recombination�seed�polish�callback�disp�init)�DifferentialEvolutionSolver�solve)�func�boundsr r r r r r r r r r r r Zsolver� r �/_differentialevolution.pyr s � c @ s� e Z d Z d Z d d d d d d d d d d i Z d d d d d d d d d d i Z f d d d d dC d d e j d d d d d d � Z d d � Z d d � Z e d! d" � � Z e d# d$ � � Z d% d& � Z d' d( � Z d) d* � Z d+ d, � Z d- d. � Z d/ d0 � Z d1 d2 � Z d3 d4 � Z d5 d6 � Z d7 d8 � Z d9 d: � Z d; d< � Z d= d> � Z d? d@ � Z dA dB � Z d S)Dr a This class implements the differential evolution solver Parameters ---------- func : callable The objective function to be minimized. Must be in the form ``f(x, *args)``, where ``x`` is the argument in the form of a 1-D array and ``args`` is a tuple of any additional fixed parameters needed to completely specify the function. bounds : sequence Bounds for variables. ``(min, max)`` pairs for each element in ``x``, defining the lower and upper bounds for the optimizing argument of `func`. It is required to have ``len(bounds) == len(x)``. ``len(bounds)`` is used to determine the number of parameters in ``x``. args : tuple, optional Any additional fixed parameters needed to completely specify the objective function. strategy : str, optional The differential evolution strategy to use. Should be one of: - 'best1bin' - 'best1exp' - 'rand1exp' - 'randtobest1exp' - 'best2exp' - 'rand2exp' - 'randtobest1bin' - 'best2bin' - 'rand2bin' - 'rand1bin' The default is 'best1bin' maxiter : int, optional The maximum number of generations over which the entire population is evolved. The maximum number of function evaluations (with no polishing) is: ``(maxiter + 1) * popsize * len(x)`` popsize : int, optional A multiplier for setting the total population size. The population has ``popsize * len(x)`` individuals. tol : float, optional When the mean of the population energies, multiplied by tol, divided by the standard deviation of the population energies is greater than 1 the solving process terminates: ``convergence = mean(pop) * tol / stdev(pop) > 1`` mutation : float or tuple(float, float), optional The mutation constant. In the literature this is also known as differential weight, being denoted by F. If specified as a float it should be in the range [0, 2]. If specified as a tuple ``(min, max)`` dithering is employed. Dithering randomly changes the mutation constant on a generation by generation basis. The mutation constant for that generation is taken from U[min, max). Dithering can help speed convergence significantly. Increasing the mutation constant increases the search radius, but will slow down convergence. recombination : float, optional The recombination constant, should be in the range [0, 1]. In the literature this is also known as the crossover probability, being denoted by CR. Increasing this value allows a larger number of mutants to progress into the next generation, but at the risk of population stability. seed : int or `np.random.RandomState`, optional If `seed` is not specified the `np.random.RandomState` singleton is used. If `seed` is an int, a new `np.random.RandomState` instance is used, seeded with `seed`. If `seed` is already a `np.random.RandomState` instance, then that `np.random.RandomState` instance is used. Specify `seed` for repeatable minimizations. disp : bool, optional Display status messages callback : callable, `callback(xk, convergence=val)`, optional A function to follow the progress of the minimization. ``xk`` is the current value of ``x0``. ``val`` represents the fractional value of the population convergence. When ``val`` is greater than one the function halts. If callback returns `True`, then the minimization is halted (any polishing is still carried out). polish : bool, optional If True, then `scipy.optimize.minimize` with the `L-BFGS-B` method is used to polish the best population member at the end. This requires a few more function evaluations. maxfun : int, optional Set the maximum number of function evaluations. However, it probably makes more sense to set `maxiter` instead. init : string, optional Specify which type of population initialization is performed. Should be one of: - 'latinhypercube' - 'random' r �_best1�randtobest1bin� _randtobest1Zbest2bin�_best2Zrand2bin�_rand2Zrand1bin�_rand1Zbest1expZrand1exp�randtobest1expZbest2expZrand2expi� r g{�G�z�?� �?r gffffff�?NFTr c C s� | | j k r+ t | | j | � | _ n7 | | j k rV t | | j | � | _ n t d � � | | _ | | _ | | _ | | _ | | _ t j t j | � � s� t j t j | � d k � s� t j t j | � d k � r� t d � � d | _ t | d � r>t | � d k r>| d | d g | _ | j j � | | _ | | _ | | _ t j | d d �j | _ t j | j d � d k s�t j t j | j � � r�t d � � | d k r�d } | | _ | d k r�t j } | | _ d | j d | j d | _ t j | j d | j d � | _ t j | j d � | _ t | � | _! | | j | _" | j" | j f | _# d | _$ | d k r�| j% � n% | d k r�| j& � n t d � � | | _' d S)Nz'Please select a valid mutation strategy� r z�The mutation constant must be a float in U[0, 2), or specified as a tuple(min, max) where min < max and min, max are in U[0, 2).�__iter__r Zdtype�floatzWbounds should be a sequence containing real valued (min, max) pairs for each value in xi� g �?r �randomzOThe population initialization method must be oneof 'latinhypercube' or 'random')(� _binomial�getattr� mutation_func� _exponential� ValueErrorr r r r �scale�np�all�isfinite�any�array�dither�hasattr�len�sort�cross_over_probabilityr r �T�limits�sizer �inf�maxfun�(_DifferentialEvolutionSolver__scale_arg1�fabs�(_DifferentialEvolutionSolver__scale_arg2�parameter_count�_make_random_gen�random_number_generator�num_population_members�population_shape�_nfev�init_population_lhs�init_population_randomr )�selfr r r r r r r r r r r? r r r r r r r �__init__; sZ ! $ z$DifferentialEvolutionSolver.__init__c C s� | j } d | j } | | j | j � t j d d | j d d �d d � t j f } t j | � | _ xR t | j � D]A } | j t | j � � } | | | f | j d d � | f <q} Wt j | j � t j | _ d | _ d S)z� Initializes the population with Latin Hypercube Sampling. Latin Hypercube Sampling ensures that each parameter is uniformly sampled over its range. g �?g ZendpointFNr )rE rF � random_samplerG r1 ZlinspaceZnewaxisZ zeros_like� population�rangerC Z permutation�onesr> �population_energiesrH )rK �rngZsegsize�samples�j�orderr r r rI � s ' z/DifferentialEvolutionSolver.init_population_lhsc C sG | j } | j | j � | _ t j | j � t j | _ d | _ d S)z� Initialises the population at random. This type of initialization can possess clustering, Latin Hypercube sampling is generally better. r N) rE rM rG rN r1 rP rF r> rQ rH )rK rR r r r rJ � s z2DifferentialEvolutionSolver.init_population_randomc C s | j | j d � S)z� The best solution from the solver Returns ------- x : ndarray The best solution from the solver. r )�_scale_parametersrN )rK r r r �x� s z DifferentialEvolutionSolver.xc C s- t j | j � t j t j | j � t � S)zb The standard deviation of the population energies divided by their mean. )r1 ZstdrQ �absZmean�_MACHEPS)rK r r r � convergence� s z'DifferentialEvolutionSolver.convergencec C s. d \ } } t d } t j t j | j � � r; | j � x� t d | j d � D]� } y t | � Wn# t k r� d } t d } PYn X| j r� t d | | j d f � | j } | j r| j | j | j d � d | j | �d k rd } d } P| | j k s| rR PqR Wt d } d } t d | j d | j d d | j d | d | d | d k � } | j r*t | j t j | j � d d d | j j d | j �} | j | j 7_ | j | _ | j | j k r*| j | _ | j | _ | j | _ | j | j d <| j | j � | j d <| S)a� Runs the DifferentialEvolutionSolver. Returns ------- res : OptimizeResult The optimization result represented as a ``OptimizeResult`` object. Important attributes are: ``x`` the solution array, ``success`` a Boolean flag indicating if the optimizer exited successfully and ``message`` which describes the cause of the termination. See `OptimizeResult` for a description of other attributes. If `polish` was employed, and a lower minimum was obtained by the polishing, then OptimizeResult also contains the ``jac`` attribute. r F�successr TZmaxfevz(differential_evolution step %d: f(x)= %grZ z8callback function requested stop early by returning Truer rW �fun�nfev�nit�message�methodzL-BFGS-Br r )r F) r r1 r2 �isinfrQ � _calculate_population_energiesrO r �next� StopIterationr �printrZ r rV rN r r rW rH r r r �copyr< r; r r] r\ Zjac�_unscale_parameters)rK r^ Z warning_flagZstatus_messagerZ Z DE_result�resultr r r r � s` z!DifferentialEvolutionSolver.solvec C s� xj t | j � D]Y \ } } | j | j k r/ P| j | � } | j | | j � | j | <| j d 7_ q Wt j | j � } | j | } | j d | j | <| | j d <| j | d g d d � f | j d | g d d � f <d S)z� Calculate the energies of all the population members at the same time. Puts the best member in first place. Useful if the population has just been initialised. r r N) � enumeraterN rH r? rV r r rQ r1 Zargmin)rK �index� candidate� parametersZminvalZ lowest_energyr r r rb 2 s z:DifferentialEvolutionSolver._calculate_population_energiesc C s | S)Nr )rK r r r r( J s z$DifferentialEvolutionSolver.__iter__c C sV t j t j | j � � r% | j � | j d k rg | j j � | j d | j d | j d | _ x� t | j � D]� } | j | j k r� t � | j | � } | j | � | j | � } | j | | j � } | j d 7_ | | j | k rw | | j | <| | j | <| | j d k rw | | j d <| | j d <qw W| j | j d f S)z� Evolve the population by a single generation Returns ------- x : ndarray The best solution from the solver. fun : float Value of objective function obtained from the best solution. Nr r )r1 r2 ra rQ rb r6 rE �randr0 rO rF rH r? rd �_mutate�_ensure_constraintrV r r rN rW )rK rk �trialrl Zenergyr r r �__next__M s( ' z$DifferentialEvolutionSolver.__next__c C s | j � S)z� Evolve the population by a single generation Returns ------- x : ndarray The best solution from the solver. fun : float Value of objective function obtained from the best solution. )rq )rK r r r rc � s z DifferentialEvolutionSolver.nextc C s | j | d | j S)zD scale from a number between 0 and 1 to parameters. g �?)r@ rB )rK rp r r r rV � s z-DifferentialEvolutionSolver._scale_parametersc C s | | j | j d S)zD scale from parameters to a number between 0 and 1. g �?)r@ rB )rK rl r r r rg � s z/DifferentialEvolutionSolver._unscale_parametersc C sL xE t | � D]7 \ } } | d k s1 | d k r | j j � | | <q Wd S)zA make sure the parameters lie between the limits r r N)ri rE rm )rK rp rj Zparamr r r ro � s z.DifferentialEvolutionSolver._ensure_constraintc C sX t j | j | � } | j } | j d | j � } | j d k sR | j d k rs | j | | j | d � � } n | j | j | d � � } | j | j k r� | j | j � } | | j k } d | | <t j | | | � } | S| j | j k rTd } xQ | | j k rO| j � | j k rO| | | | <| d | j } | d 7} q� W| Sd S)zD create a trial vector based on a mutation strategy r r% r � Tr N)r1 rf rN rE �randintrC r r- �_select_samplesr+ rm r: Zwherer. )rK rk rp rR Z fill_point�bprimeZ crossovers�ir r r rn � s, z#DifferentialEvolutionSolver._mutatec C s> | d d � \ } } | j d | j | j | | j | S)z$ best1bin, best1exp Nr' r )rN r0 )rK rS �r0�r1r r r r � s z"DifferentialEvolutionSolver._best1c C sA | d d � \ } } } | j | | j | j | | j | S)z$ rand1bin, rand1exp N� )rN r0 )rK rS rw rx �r2r r r r$ � s z"DifferentialEvolutionSolver._rand1c C so | d d � \ } } t j | j | � } | | j | j d | 7} | | j | j | | j | 7} | S)z0 randtobest1bin, randtobest1exp Nr' r )r1 rf rN r0 )rK rk rS rw rx ru r r r r! � s z(DifferentialEvolutionSolver._randtobest1c C s` | d d � \ } } } } | j d | j | j | | j | | j | | j | } | S)z$ best2bin, best2exp N� r )rN r0 )rK rS rw rx rz �r3ru r r r r" � s 0z"DifferentialEvolutionSolver._best2c C sY | \ } } } } } | j | | j | j | | j | | j | | j | } | S)z$ rand2bin, rand2exp )rN r0 )rK rS rw rx rz r| Zr4ru r r r r# � s 0z"DifferentialEvolutionSolver._rand2c C sF t t | j � � } | j | � | j j | � | d | � } | S)z� obtain random integers from range(self.num_population_members), without replacement. You can't have the original candidate either. N)�listrO rF �removerE �shuffle)rK rk Znumber_samplesZidxsr r r rt � s z+DifferentialEvolutionSolver._select_samples)r& r ) �__name__� __module__� __qualname__�__doc__r+ r. r1 r> rL rI rJ �propertyrW rZ r rb r( rq rc rV rg ro rn r r$ r! r" r# rt r r r r r � sF \ P & Y 3 ' r c C s� | d k s | t j k r( t j j j St | t j t j f � rS t j j | � St | t j j � rl | St d | � � d S)a# Turn seed into a np.random.RandomState instance If seed is None, return the RandomState singleton used by np.random. If seed is an int, return a new RandomState instance seeded with seed. If seed is already a RandomState instance, return it. Otherwise raise ValueError. Nz=%r cannot be used to seed a numpy.random.RandomState instance) r1 r* ZmtrandZ_rand� isinstance�numbersZIntegralZintegerZ RandomStater/ )r r r r rD s rD )r� Z __future__r r r Znumpyr1 Zscipy.optimizer r Zscipy.optimize.optimizer r� �__all__ZfinfoZfloat64ZepsrY r �objectr rD r r r r �<module> s �� � 7
Submit
FILE
FOLDER
Name
Size
Permission
Action
__init__.cpython-35.opt-1.pyc
6964 bytes
0644
__init__.cpython-35.pyc
6964 bytes
0644
_basinhopping.cpython-35.opt-1.pyc
24677 bytes
0644
_basinhopping.cpython-35.pyc
24677 bytes
0644
_differentialevolution.cpython-35.opt-1.pyc
26738 bytes
0644
_differentialevolution.cpython-35.pyc
26738 bytes
0644
_hungarian.cpython-35.opt-1.pyc
7804 bytes
0644
_hungarian.cpython-35.pyc
7804 bytes
0644
_linprog.cpython-35.opt-1.pyc
31172 bytes
0644
_linprog.cpython-35.pyc
31172 bytes
0644
_minimize.cpython-35.opt-1.pyc
24043 bytes
0644
_minimize.cpython-35.pyc
24043 bytes
0644
_numdiff.cpython-35.opt-1.pyc
18899 bytes
0644
_numdiff.cpython-35.pyc
18899 bytes
0644
_root.cpython-35.opt-1.pyc
25483 bytes
0644
_root.cpython-35.pyc
25483 bytes
0644
_spectral.cpython-35.opt-1.pyc
7761 bytes
0644
_spectral.cpython-35.pyc
7761 bytes
0644
_trustregion.cpython-35.opt-1.pyc
7169 bytes
0644
_trustregion.cpython-35.pyc
7169 bytes
0644
_trustregion_dogleg.cpython-35.opt-1.pyc
3735 bytes
0644
_trustregion_dogleg.cpython-35.pyc
3735 bytes
0644
_trustregion_ncg.cpython-35.opt-1.pyc
3464 bytes
0644
_trustregion_ncg.cpython-35.pyc
3464 bytes
0644
_tstutils.cpython-35.opt-1.pyc
1995 bytes
0644
_tstutils.cpython-35.pyc
1995 bytes
0644
cobyla.cpython-35.opt-1.pyc
9239 bytes
0644
cobyla.cpython-35.pyc
9239 bytes
0644
lbfgsb.cpython-35.opt-1.pyc
15952 bytes
0644
lbfgsb.cpython-35.pyc
15952 bytes
0644
linesearch.cpython-35.opt-1.pyc
18967 bytes
0644
linesearch.cpython-35.pyc
18967 bytes
0644
minpack.cpython-35.opt-1.pyc
28403 bytes
0644
minpack.cpython-35.pyc
28403 bytes
0644
nnls.cpython-35.opt-1.pyc
1659 bytes
0644
nnls.cpython-35.pyc
1659 bytes
0644
nonlin.cpython-35.opt-1.pyc
47479 bytes
0644
nonlin.cpython-35.pyc
47545 bytes
0644
optimize.cpython-35.opt-1.pyc
83901 bytes
0644
optimize.cpython-35.pyc
83901 bytes
0644
setup.cpython-35.opt-1.pyc
2801 bytes
0644
setup.cpython-35.pyc
2801 bytes
0644
slsqp.cpython-35.opt-1.pyc
15484 bytes
0644
slsqp.cpython-35.pyc
15484 bytes
0644
tnc.cpython-35.opt-1.pyc
13998 bytes
0644
tnc.cpython-35.pyc
13998 bytes
0644
zeros.cpython-35.opt-1.pyc
19335 bytes
0644
zeros.cpython-35.pyc
19335 bytes
0644
N4ST4R_ID | Naxtarrr
