sfepy.solvers.optimize module¶

class sfepy.solvers.optimize.FMinSteepestDescent(conf, **kwargs)[source]¶

Steepest descent optimization solver.

Kind: ‘opt.fmin_sd’

For common configuration parameters, see Solver.

Specific configuration parameters:

Parameters:

i_maxint (default: 10): The maximum number of iterations.
eps_rdfloat (default: 1e-05): The relative delta of the objective function.
eps_offloat (default: 0.0001): The tolerance for the objective function.
eps_ofgfloat (default: 1e-08): The tolerance for the objective function gradient.
normnumpy norm (default: inf): The norm to be used.
lsbool (default: True): If True, use a line-search.
ls_method{‘backtracking’, ‘full’} (default: ‘backtracking’): The line-search method.
ls_onfloat (default: 0.99999): Start the backtracking line-search by reducing the step, if $||f(x^i)|| / ||f(x^{i-1})||$ is larger than ls_on.
ls00.0 < float < 1.0 (default: 1.0): The initial step.
ls_red0.0 < float < 1.0 (default: 0.5): The step reduction factor in case of correct residual assembling.
ls_red_warp0.0 < float < 1.0 (default: 0.1): The step reduction factor in case of failed residual assembling (e.g. the “warp violation” error caused by a negative volume element resulting from too large deformations).
ls_min0.0 < float < 1.0 (default: 1e-05): The minimum step reduction factor.
check0, 1 or 2 (default: 0): If >= 1, check the tangent matrix using finite differences. If 2, plot the resulting sparsity patterns.
deltafloat (default: 1e-06): If check >= 1, the finite difference matrix is taken as $A_{ij} = \frac{f_i(x_j + \delta) - f_i(x_j - \delta)}{2 \delta}$ .
outputfunction: If given, use it instead of output() function.
yscaleslist of str (default: [‘linear’, ‘log’, ‘log’, ‘linear’]): The list of four convergence log subplot scales.
logdict or None: If not None, log the convergence according to the configuration in the following form: {'text' : 'log.txt', 'plot' : 'log.pdf'}. Each of the dict items can be None.

name = 'opt.fmin_sd'¶

class sfepy.solvers.optimize.ScipyFMinSolver(conf, **kwargs)[source]¶

Interface to SciPy optimization solvers scipy.optimize.fmin_*.

Kind: ‘nls.scipy_fmin_like’

For common configuration parameters, see Solver.

Specific configuration parameters:

Parameters:

method{‘fmin’, ‘fmin_bfgs’, ‘fmin_cg’, ‘fmin_cobyla’, ‘fmin_l_bfgs_b’, ‘fmin_ncg’, ‘fmin_powell’, ‘fmin_slsqp’, ‘fmin_tnc’} (default: ‘fmin’): The actual optimization method to use.
i_maxint (default: 10): The maximum number of iterations.
**: Additional parameters supported by the method.

name = 'nls.scipy_fmin_like'¶

set_method(conf)[source]¶

sfepy.solvers.optimize.check_gradient(xit, aofg, fn_of, delta, check)[source]¶

sfepy.solvers.optimize.conv_test(conf, it, of, of0, ofg_norm=None)[source]¶

Returns:

flagint

-1 … continue
0 … small OF -> stop
1 … i_max reached -> stop
2 … small OFG -> stop
3 … small relative decrase of OF

sfepy.solvers.optimize.wrap_function(function, args)[source]¶