scipy_optimizer.py

OpenMDAO Wrapper for the scipy.optimize.minimize family of local optimizers.

class openmdao.drivers.scipy_optimizer.ScipyOptimizeDriver(**kwargs)

Bases: Driver

Driver wrapper for the scipy.optimize.minimize family of local optimizers.

Inequality constraints are supported by COBYLA and SLSQP, but equality constraints are only supported by SLSQP. None of the other optimizers support constraints.

ScipyOptimizeDriver supports the following:

equality_constraints inequality_constraints

Parameters:

**kwargsdict of keyword arguments

Keyword arguments that will be mapped into the Driver options.

Attributes:

failbool

Flag that indicates failure of most recent optimization.

iter_countint

Counter for function evaluations.

_scipy_optimize_resultOptimizeResult

Result returned from scipy.optimize call.

opt_settingsdict

Dictionary of solver-specific options. See the scipy.optimize.minimize documentation.

_check_jacbool

Used internally to control when to perform singular checks on computed total derivs.

_con_cachedict

Cached result of constraint evaluations because scipy asks for them in a separate function.

_con_idxdict

Used for constraint bookkeeping in the presence of 2-sided constraints.

_grad_cache{}

Cached result of nonlinear constraint derivatives because scipy asks for them in a separate function.

_exc_info3 item tuple

Storage for exception and traceback information.

_obj_and_nlconslist

List of objective + nonlinear constraints. Used to compute total derivatives for all except linear constraints.

_dvlistlist

Copy of _designvars.

_lincongrad_cachenp.ndarray

Pre-calculated gradients of linear constraints.

_desvar_array_cachenp.ndarray

Cached array for setting design variables.

__init__(**kwargs)

Initialize the ScipyOptimizeDriver.

add_recorder(recorder)

Add a recorder to the driver.

Parameters:

recorderCaseRecorder

A recorder instance.

check_relevance()

Check if there are constraints that don’t depend on any design vars.

This usually indicates something is wrong with the problem formulation.

cleanup()

Clean up resources prior to exit.

declare_coloring(num_full_jacs=3, tol=1e-25, orders=None, perturb_size=1e-09, min_improve_pct=5.0, show_summary=True, show_sparsity=False, use_scaling=False, randomize_subjacs=True, randomize_seeds=False, direct=True)

Set options for total deriv coloring.

Parameters:

num_full_jacsint

Number of times to repeat partial jacobian computation when computing sparsity.

tolfloat

Tolerance used to determine if an array entry is nonzero during sparsity determination.

ordersint

Number of orders above and below the tolerance to check during the tolerance sweep.

perturb_sizefloat

Size of input/output perturbation during generation of sparsity.

min_improve_pctfloat

If coloring does not improve (decrease) the number of solves more than the given percentage, coloring will not be used.

show_summarybool

If True, display summary information after generating coloring.

show_sparsitybool

If True, display sparsity with coloring info after generating coloring.

use_scalingbool

If True, use driver scaling when generating the sparsity.

randomize_subjacsbool

If True, use random subjacobians corresponding to their declared sparsity patterns.

randomize_seedsbool

If True, use random seeds when computing the sparsity.

directbool

If using bidirectional coloring, use the direct method when computing the column adjacency matrix instead of the substitution method.

get_constraint_values(ctype='all', lintype='all', driver_scaling=True)

Return constraint values.

Parameters:

ctypestr

Default is ‘all’. Optionally return just the inequality constraints with ‘ineq’ or the equality constraints with ‘eq’.

lintypestr

Default is ‘all’. Optionally return just the linear constraints with ‘linear’ or the nonlinear constraints with ‘nonlinear’.

driver_scalingbool

When True, return values that are scaled according to either the adder and scaler or the ref and ref0 values that were specified when add_design_var, add_objective, and add_constraint were called on the model. Default is True.

Returns:

dict

Dictionary containing values of each constraint.

get_design_var_values(get_remote=True, driver_scaling=True)

Return the design variable values.

Parameters:

get_remotebool or None

If True, retrieve the value even if it is on a remote process. Note that if the variable is remote on ANY process, this function must be called on EVERY process in the Problem’s MPI communicator. If False, only retrieve the value if it is on the current process, or only the part of the value that’s on the current process for a distributed variable.

driver_scalingbool

When True, return values that are scaled according to either the adder and scaler or the ref and ref0 values that were specified when add_design_var, add_objective, and add_constraint were called on the model. Default is True.

Returns:

dict

Dictionary containing values of each design variable.

get_driver_derivative_calls()

Return number of derivative evaluations made during a driver run.

Returns:

int

Number of derivative evaluations made during a driver run.

get_driver_objective_calls()

Return number of objective evaluations made during a driver run.

Returns:

int

Number of objective evaluations made during a driver run.

get_exit_status()

Return exit status of driver run.

Returns:

str

String indicating result of driver run.

get_objective_values(driver_scaling=True)

Return objective values.

Parameters:

driver_scalingbool

When True, return values that are scaled according to either the adder and scaler or the ref and ref0 values that were specified when add_design_var, add_objective, and add_constraint were called on the model. Default is True.

Returns:

dict

Dictionary containing values of each objective.

get_reports_dir()

Get the path to the directory where the report files should go.

If it doesn’t exist, it will be created.

Returns:

str

The path to the directory where reports should be written.

property msginfo

Return info to prepend to messages.

Returns:

str

Info to prepend to messages.

record_derivatives()

Record the current total jacobian.

record_iteration()

Record an iteration of the current Driver.

run()

Optimize the problem using selected Scipy optimizer.

Returns:

bool

Failure flag; True if failed to converge, False is successful.

scaling_report(outfile='driver_scaling_report.html', title=None, show_browser=True, jac=True)

Generate a self-contained html file containing a detailed connection viewer.

Optionally pops up a web browser to view the file.

Parameters:

outfilestr, optional

The name of the output html file. Defaults to ‘driver_scaling_report.html’.

titlestr, optional

Sets the title of the web page.

show_browserbool, optional

If True, pop up a browser to view the generated html file. Defaults to True.

jacbool

If True, show jacobian information.

Returns:

dict

Data used to create html file.

set_design_var(name, value, set_remote=True)

Set the value of a design variable.

‘name’ can be a promoted output name or an alias.

Parameters:

namestr

Global pathname of the design variable.

valuefloat or ndarray

Value for the design variable.

set_remotebool

If True, set the global value of the variable (value must be of the global size). If False, set the local value of the variable (value must be of the local size).

use_fixed_coloring(coloring=<object object>)

Tell the driver to use a precomputed coloring.

Parameters:

coloringstr or Coloring

A coloring filename or a Coloring object. If no arg is passed, filename will be determined automatically.

openmdao.drivers.scipy_optimizer.signature_extender(fcn, extra_args)

Closure function, which appends extra arguments to the original function call.

The first argument is the design vector. The possible extra arguments from the callback of scipy.optimize.minimize() are not passed to the function.

Some algorithms take a sequence of NonlinearConstraint as input for the constraints. For this class it is not possible to pass additional arguments. With this function the signature will be correct for both scipy and the driver.

Parameters:

fcncallable

Function, which takes the design vector as the first argument.

extra_argstuple or list

Extra arguments for the function.

Returns:

callable

The function with the signature expected by the driver.