"""Define the Component class."""
import sys
import inspect
from collections import defaultdict
from collections.abc import Iterable
from itertools import product, chain
from io import StringIO
from numbers import Integral
import numpy as np
from numpy import ndarray, isscalar, ndim, atleast_1d, atleast_2d, promote_types
from scipy.sparse import issparse, coo_matrix, csr_matrix
from openmdao.core.system import System, _supported_methods, _DEFAULT_COLORING_META, \
global_meta_names, collect_errors, _iter_derivs
from openmdao.core.constants import INT_DTYPE, _DEFAULT_OUT_STREAM, _SetupStatus
from openmdao.jacobians.dictionary_jacobian import _CheckingJacobian
from openmdao.utils.units import simplify_unit
from openmdao.utils.name_maps import abs_key_iter, abs_key2rel_key, rel_key2abs_key
from openmdao.utils.mpi import MPI
from openmdao.utils.array_utils import shape_to_len, submat_sparsity_iter, sparsity_diff_viz
from openmdao.utils.deriv_display import _deriv_display, _deriv_display_compact
from openmdao.utils.general_utils import format_as_float_or_array, ensure_compatible, \
find_matches, make_set, inconsistent_across_procs, LocalRangeIterable
from openmdao.utils.indexer import Indexer, indexer
import openmdao.utils.coloring as coloring_mod
from openmdao.utils.om_warnings import issue_warning, MPIWarning, DistributedComponentWarning, \
DerivativesWarning, warn_deprecation, OMInvalidCheckDerivativesOptionsWarning
from openmdao.utils.code_utils import is_lambda, LambdaPickleWrapper, get_function_deps, \
get_return_names
from openmdao.approximation_schemes.complex_step import ComplexStep
from openmdao.approximation_schemes.finite_difference import FiniteDifference
_forbidden_chars = {'.', '*', '?', '!', '[', ']'}
_whitespace = {' ', '\t', '\r', '\n'}
_allowed_types = (list, tuple, ndarray, Iterable)
def _valid_var_name(name):
"""
Determine if the proposed name is a valid variable name.
Leading and trailing whitespace is illegal, and a specific list of characters
are illegal anywhere in the string.
Parameters
----------
name : str
Proposed name.
Returns
-------
bool
True if the proposed name is a valid variable name, else False.
"""
global _forbidden_chars, _whitespace
if not name:
return False
if _forbidden_chars.intersection(name):
return False
return name is name.strip()
[docs]class Component(System):
"""
Base Component class; not to be directly instantiated.
Parameters
----------
**kwargs : dict of keyword arguments
Available here and in all descendants of this system.
Attributes
----------
_var_rel2meta : dict
Dictionary mapping relative names to metadata.
This is only needed while adding inputs and outputs. During setup, these are used to
build the dictionaries of metadata. This contains both continuous and discrete variables.
_static_var_rel2meta : dict
Static version of above - stores data for variables added outside of setup.
_var_rel_names : {'input': [str, ...], 'output': [str, ...]}
List of relative names of owned variables existing on current proc.
This is only needed while adding inputs and outputs. During setup, these are used to
determine the list of absolute names. This includes only continuous variables.
_static_var_rel_names : dict
Static version of above - stores names of variables added outside of setup.
_declared_partials_patterns : dict
Dictionary of declared partials patterns. Each key is a tuple of the form
(of, wrt) where of and wrt may be glob patterns.
_declared_partial_checks : list
Cached storage of user-declared check partial options.
_no_check_partials : bool
If True, the check_partials function will ignore this component.
_has_distrib_outputs : bool
If True, this component has at least one distributed output.
_compute_primals_out_shape : tuple or None
Cached (shape, istuple) of the output from compute_primal function. If istuple is True,
then shape is a tuple of shapes, otherwise it is a single shape.
_valid_name_map : dict
Mapping of declared input/output names to valid Python names.
_orig_compute_primal : function
The original compute_primal method.
"""
[docs] def __init__(self, **kwargs):
"""
Initialize all attributes.
"""
super().__init__(**kwargs)
self._var_rel_names = {'input': [], 'output': []}
self._var_rel2meta = {}
self._static_var_rel_names = {'input': [], 'output': []}
self._static_var_rel2meta = {}
self._declared_partials_patterns = {}
self._declared_partial_checks = []
self._no_check_partials = False
self._has_distrib_outputs = False
self._compute_primals_out_shape = None
self._valid_name_map = {}
self._orig_compute_primal = getattr(self, 'compute_primal')
def _tree_flatten(self):
"""
Return a flattened pytree representation of this component.
We treat this component, when passed as 'self' into a function that is used by jax, as a
pytree with no continuous data.
Returns
-------
Tuple
A tuple containing continuous and static data.
"""
return ((), {'_self_': self, '_statics_': self.get_self_statics()})
@staticmethod
def _tree_unflatten(aux_data, children):
"""
Return the same instance of this component that was returned by the _tree_flatten method.
"""
return aux_data['_self_']
def _declare_options(self):
"""
Declare options before kwargs are processed in the init method.
"""
super()._declare_options()
self.options.declare('distributed', types=bool, default=False,
desc='If True, set all variables in this component as distributed '
'across multiple processes')
self.options.declare('run_root_only', types=bool, default=False,
desc='If True, call compute, compute_partials, linearize, '
'apply_linear, apply_nonlinear, solve_linear, solve_nonlinear, '
'and compute_jacvec_product only on rank 0 and broadcast the '
'results to the other ranks.')
self.options.declare('always_opt', types=bool, default=False,
desc='If True, force nonlinear operations on this component to be '
'included in the optimization loop even if this component is not '
'relevant to the design variables and responses.')
self.options.declare('use_jit', types=bool, default=True,
desc='If True, attempt to use jit on compute_primal, assuming jax or '
'some other AD package capable of jitting is active.')
self.options.declare('default_shape', types=tuple, default=(1,),
desc='Default shape for variables that do not set val to a non-scalar '
'value or set shape, shape_by_conn, copy_shape, or compute_shape.'
' Default is (1,).')
[docs] def setup(self):
"""
Declare inputs and outputs.
Available attributes:
name
pathname
comm
options
"""
pass
def _setup_procs(self, pathname, comm, prob_meta):
"""
Execute first phase of the setup process.
Distribute processors, assign pathnames, and call setup on the component.
Parameters
----------
pathname : str
Global name of the system, including the path.
comm : MPI.Comm or <FakeComm>
MPI communicator object.
prob_meta : dict
Problem level metadata.
"""
super()._setup_procs(pathname, comm, prob_meta)
if self._num_par_fd > 1:
if comm.size > 1:
comm = self._setup_par_fd_procs(comm)
elif not MPI:
issue_warning(f"MPI is not active but num_par_fd = {self._num_par_fd}. No parallel "
"finite difference will be performed.",
prefix=self.msginfo, category=MPIWarning)
self._num_par_fd = 1
self.comm = comm
# Clear out old variable information so that we can call setup on the component.
self._var_rel_names = {'input': [], 'output': []}
self._var_rel2meta = {}
if comm.size == 1:
self._has_distrib_vars = self._has_distrib_outputs = False
for meta in self._static_var_rel2meta.values():
# reset shape if any dynamic shape parameters are set in case this is a resetup
# NOTE: this is necessary because we allow variables to be added in __init__.
if 'shape_by_conn' in meta and (meta['shape_by_conn'] or
meta['compute_shape'] is not None):
meta['shape'] = None
if not isscalar(meta['val']):
if meta['val'].size > 0:
meta['val'] = meta['val'].flatten()[0]
else:
meta['val'] = 1.0
self._var_rel2meta.update(self._static_var_rel2meta)
for io in ['input', 'output']:
self._var_rel_names[io].extend(self._static_var_rel_names[io])
self.setup()
self._setup_check()
self._set_vector_class()
def _set_vector_class(self):
if self._has_distrib_vars:
dist_vec_class = self._problem_meta['distributed_vector_class']
if dist_vec_class is not None:
self._vector_class = dist_vec_class
else:
issue_warning("Component contains distributed variables, "
"but there is no distributed vector implementation (MPI/PETSc) "
"available. The default non-distributed vectors will be used.",
prefix=self.msginfo, category=DistributedComponentWarning)
self._vector_class = self._problem_meta['local_vector_class']
else:
self._vector_class = self._problem_meta['local_vector_class']
def _configure_check(self):
"""
Do any error checking on i/o configuration.
"""
# Check here if declare_coloring was called during setup but declare_partials wasn't.
# If declare partials wasn't called, call it with of='*' and wrt='*' so we'll have
# something to color.
if self._coloring_info.coloring is not None:
for meta in self._declared_partials_patterns.values():
if 'method' in meta and meta['method'] is not None:
break
else:
method = self._coloring_info.method
issue_warning("declare_coloring or use_fixed_coloring was called but no approx"
" partials were declared. Declaring all partials as approximated "
f"using default metadata and method='{method}'.", prefix=self.msginfo,
category=DerivativesWarning)
self.declare_partials('*', '*', method=method)
super()._configure_check()
def _setup_var_data(self):
"""
Compute the list of abs var names, abs/prom name maps, and metadata dictionaries.
"""
global global_meta_names
super()._setup_var_data()
# Compute the prefix for turning rel/prom names into abs names
prefix = self.pathname + '.'
for io in ['input', 'output']:
abs2meta = self._var_abs2meta[io]
allprocs_abs2meta = self._var_allprocs_abs2meta[io]
is_input = io == 'input'
for prom_name in self._var_rel_names[io]:
abs_name = prefix + prom_name
abs2meta[abs_name] = metadata = self._var_rel2meta[prom_name]
self._resolver.add_mapping(abs_name, prom_name, io,
local=True, distributed=metadata['distributed'])
allprocs_abs2meta[abs_name] = {
meta_name: metadata[meta_name]
for meta_name in global_meta_names[io]
}
if is_input and 'src_indices' in metadata:
allprocs_abs2meta[abs_name]['has_src_indices'] = \
metadata['src_indices'] is not None
for prom_name, val in self._var_discrete[io].items():
abs_name = prefix + prom_name
self._resolver.add_mapping(abs_name, prom_name, io,
local=True, continuous=False)
# Compute allprocs_discrete (metadata for discrete vars)
self._var_allprocs_discrete[io][abs_name] = v = val.copy()
del v['val']
if self._var_discrete['input'] or self._var_discrete['output']:
self._discrete_inputs = _DictValues(self._var_discrete['input'])
self._discrete_outputs = _DictValues(self._var_discrete['output'])
else:
self._discrete_inputs = self._discrete_outputs = {}
if self.comm.size > 1:
# check that same variables are declared on all procs
vnames = (list(self._var_rel_names['output']), list(self._var_rel_names['input']))
allnames = self.comm.gather(vnames, root=0)
if self.comm.rank == 0:
outset, inset = vnames
msg = ''
for oset, iset in allnames:
if iset != inset or oset != outset:
msg = self._missing_vars_error(allnames)
break
self.comm.bcast(msg, root=0)
else:
msg = self.comm.bcast(None, root=0)
if msg:
raise RuntimeError(msg)
if self.compute_primal is not None:
self._check_compute_primal_args()
self._check_compute_primal_returns()
self._serial_idxs = None
self._inconsistent_keys = set()
def _missing_vars_error(self, allnames):
msg = ''
outset, inset = allnames[0]
for rank, (olist, ilist) in enumerate(allnames):
if rank != 0 and (olist != outset or ilist != inset):
idiff = set(inset).symmetric_difference(ilist)
odiff = set(outset).symmetric_difference(olist)
if idiff or odiff:
varnames = sorted(idiff | odiff)
if len(varnames) == 1:
varmsg = f"Variable '{varnames[0]}' exists on some ranks and not others."
else:
varmsg = f"Variables {varnames} exist on some ranks and not others."
else:
varmsg = "Variables have not been declared in the same order on all ranks."
msg = (f"{self.msginfo}: {varmsg} A component must declare all variables in "
"the same order on all ranks, even if the size of the variable is 0 on "
"some ranks.")
break
return msg
@collect_errors
def _setup_var_sizes(self):
"""
Compute the arrays of variable sizes for all variables/procs on this system.
"""
iproc = self.comm.rank
abs2idx = self._var_allprocs_abs2idx = {}
for io in ('input', 'output'):
sizes = self._var_sizes[io] = np.zeros((self.comm.size, len(self._var_rel_names[io])),
dtype=INT_DTYPE)
for i, (name, metadata) in enumerate(self._var_allprocs_abs2meta[io].items()):
sz = metadata['size']
sizes[iproc, i] = 0 if sz is None else sz
abs2idx[name] = i
if self.comm.size > 1:
my_sizes = sizes[iproc, :].copy()
self.comm.Allgather(my_sizes, sizes)
self._owned_sizes = self._var_sizes['output']
def _setup_partials(self):
"""
Process all partials and approximations that the user declared.
"""
self._subjacs_info = {}
if not self.matrix_free:
self._init_jacobian()
self.setup_partials() # hook for component writers to specify sparsity patterns
if self.options['derivs_method'] in ('cs', 'fd'):
if self.matrix_free:
raise RuntimeError(f"{self.msginfo}: derivs_method of 'cs' or 'fd' is not "
"allowed for a matrix free component.")
self._has_approx = True
method = self.options['derivs_method']
self._get_approx_scheme(method)
if not self._declared_partials_patterns:
if self.compute_primal is None:
raise RuntimeError(f"{self.msginfo}: compute_primal must be defined if using "
"a derivs_method option of 'cs' or 'fd'")
# declare all partials as 'cs' or 'fd'
for of, wrt in get_function_deps(self._orig_compute_primal,
self._var_rel_names['output']):
if of in self._discrete_outputs or wrt in self._discrete_inputs:
continue
self.declare_partials(of, wrt, method=method)
else:
# declare only those partials that have been declared
for meta in self._declared_partials_patterns.values():
meta['method'] = method
# check to make sure that if num_par_fd > 1 that this system is actually doing FD.
# Unfortunately we have to do this check after system setup has been called because that's
# when declare_partials generally happens, so we raise an exception here instead of just
# resetting the value of num_par_fd (because the comm has already been split and possibly
# used by the system setup).
orig_comm = self._full_comm if self._full_comm is not None else self.comm
if self._num_par_fd > 1 and orig_comm.size > 1 and not (self._owns_approx_jac or
self._approx_schemes):
raise RuntimeError("%s: num_par_fd is > 1 but no FD is active." % self.msginfo)
for key, pattern_meta in self._declared_partials_patterns.items():
of, wrt = key
self._resolve_partials_patterns(of, wrt, pattern_meta)
[docs] def setup_partials(self):
"""
Declare partials.
This is meant to be overridden by component classes. All partials should be
declared here since this is called after all size/shape information is known for
all variables.
"""
pass
def _setup_residuals(self):
"""
Process hook to call user-defined setup_residuals method if provided.
"""
pass
def _declared_partials_iter(self):
"""
Iterate over all declared partials.
Yields
------
key : tuple (of, wrt)
Subjacobian key.
"""
yield from self._subjacs_info.keys()
def _get_missing_partials(self, missing):
"""
Provide (of, wrt) tuples for which derivatives have not been declared in the component.
Parameters
----------
missing : dict
Dictionary containing set of missing derivatives keyed by system pathname.
"""
if ('*', '*') in self._declared_partials_patterns or \
(('*',), ('*',)) in self._declared_partials_patterns:
return
# keep old default behavior where matrix free components are assumed to have
# 'dense' whole variable to whole variable partials if no partials are declared.
if self.matrix_free and not self._declared_partials_patterns:
return
keyset = self._subjacs_info
mset = set()
for of in self._var_allprocs_abs2meta['output']:
for wrt in self._var_allprocs_abs2meta['input']:
if (of, wrt) not in keyset:
mset.add((of, wrt))
if mset:
missing[self.pathname] = mset
@property
def checking(self):
"""
Return True if check_partials or check_totals is executing.
Returns
-------
bool
True if we're running within check_partials or check_totals.
"""
return self._problem_meta is not None and self._problem_meta['checking']
def _run_root_only(self):
"""
Return the value of the run_root_only option and check for possible errors.
Returns
-------
bool
True if run_root_only is active.
"""
if self.options['run_root_only']:
if self.comm.size > 1 or (self._full_comm is not None and self._full_comm.size > 1):
if self._has_distrib_vars:
raise RuntimeError(f"{self.msginfo}: Can't set 'run_root_only' option when "
"a component has distributed variables.")
if self._num_par_fd > 1:
raise RuntimeError(f"{self.msginfo}: Can't set 'run_root_only' option when "
"using parallel FD.")
if self._problem_meta['has_par_deriv_color']:
raise RuntimeError(f"{self.msginfo}: Can't set 'run_root_only' option when "
"using parallel_deriv_color.")
return True
return False
def _promoted_wrt_iter(self):
yield from self._get_partials_wrts()
[docs] def add_output(self, name, val=1.0, shape=None, units=None, res_units=None, desc='',
lower=None, upper=None, ref=1.0, ref0=0.0, res_ref=None, tags=None,
shape_by_conn=False, copy_shape=None, compute_shape=None, distributed=None,
primal_name=None):
"""
Add an output variable to the component.
Parameters
----------
name : str
Name of the variable in this component's namespace.
val : float or list or tuple or ndarray
The initial value of the variable being added in user-defined units. Default is 1.0.
shape : int or tuple or list or None
Shape of this variable, only required if val is not an array.
Default is None.
units : str or None
Units in which the output variables will be provided to the component during execution.
Default is None, which means it has no units.
res_units : str or None
Units in which the residuals of this output will be given to the user when requested.
Default is None, which means it has no units.
desc : str
Description of the variable.
lower : float or list or tuple or ndarray or Iterable or None
Lower bound(s) in user-defined units. It can be (1) a float, (2) an array_like
consistent with the shape arg (if given), or (3) an array_like matching the shape of
val, if val is array_like. A value of None means this output has no lower bound.
Default is None.
upper : float or list or tuple or ndarray or or Iterable None
Upper bound(s) in user-defined units. It can be (1) a float, (2) an array_like
consistent with the shape arg (if given), or (3) an array_like matching the shape of
val, if val is array_like. A value of None means this output has no upper bound.
Default is None.
ref : float or ndarray
Scaling parameter. The value in the user-defined units of this output variable when
the scaled value is 1. Default is 1.
ref0 : float or ndarray
Scaling parameter. The value in the user-defined units of this output variable when
the scaled value is 0. Default is 0.
res_ref : float or ndarray
Scaling parameter. The value in the user-defined res_units of this output's residual
when the scaled value is 1. Default is 1.
tags : str or list of strs or set of strs
User defined tags that can be used to filter what gets listed when calling
list_inputs and list_outputs.
shape_by_conn : bool
If True, shape this output to match its connected input(s).
copy_shape : str or None
If a str, that str is the name of a variable. Shape this output to match that of
the named variable.
compute_shape : function
A function taking a dict arg containing names and shapes of this component's inputs
and returning the shape of this output.
distributed : bool
If True, this variable is a distributed variable, so it can have different sizes/values
across MPI processes.
primal_name : str or None
Valid python name to represent the variable in compute_primal if 'name' is not a valid
python name.
Returns
-------
dict
Metadata for added variable.
"""
global _allowed_types
# First, type check all arguments
if (shape_by_conn or copy_shape or compute_shape) and (shape is not None or ndim(val) > 0):
raise ValueError("%s: If shape is to be set dynamically using 'shape_by_conn', "
"'copy_shape', or 'compute_shape', 'shape' and 'val' should be scalar,"
" but shape of '%s' and val of '%s' was given for variable '%s'."
% (self.msginfo, shape, val, name))
if not isinstance(name, str):
raise TypeError('%s: The name argument should be a string.' % self.msginfo)
if not _valid_var_name(name):
raise NameError("%s: '%s' is not a valid output name." % (self.msginfo, name))
if shape is not None and not isinstance(shape, (int, tuple, list, np.integer)):
raise TypeError("%s: The shape argument should be an int, tuple, or list but "
"a '%s' was given" % (self.msginfo, type(shape)))
if res_units is not None:
if not isinstance(res_units, str):
msg = '%s: The res_units argument should be a str or None' % self.msginfo
raise TypeError(msg)
res_units = simplify_unit(res_units, msginfo=self.msginfo)
if units is not None:
if not isinstance(units, str):
raise TypeError('%s: The units argument should be a str or None' % self.msginfo)
units = simplify_unit(units, msginfo=self.msginfo)
if tags is not None and not isinstance(tags, (str, set, list)):
raise TypeError('The tags argument should be a str, set, or list')
if not (copy_shape or shape_by_conn or compute_shape):
if not isscalar(val) and not isinstance(val, _allowed_types):
msg = '%s: The val argument should be a float, list, tuple, ndarray or Iterable'
raise TypeError(msg % self.msginfo)
default_shape = self.options['default_shape']
# value, shape: based on args, making sure they are compatible
val, shape = ensure_compatible(name, val, shape, default_shape=default_shape)
if lower is not None:
lower = ensure_compatible(name, lower, shape, default_shape=default_shape)[0]
self._has_bounds = True
if upper is not None:
upper = ensure_compatible(name, upper, shape, default_shape=default_shape)[0]
self._has_bounds = True
# All refs: check the shape if necessary
for item, item_name in zip([ref, ref0, res_ref], ['ref', 'ref0', 'res_ref']):
if item is not None and not isscalar(item):
if not isinstance(item, _allowed_types):
raise TypeError(f'{self.msginfo}: The {item_name} argument should be a '
'float, list, tuple, ndarray or Iterable')
it = atleast_1d(item)
if it.shape != shape:
raise ValueError(f"{self.msginfo}: When adding output '{name}', expected "
f"shape {shape} but got shape {it.shape} for argument "
f"'{item_name}'.")
if isscalar(ref):
self._has_output_scaling |= ref != 1.0
else:
self._has_output_scaling |= np.any(ref != 1.0)
if isscalar(ref0):
self._has_output_scaling |= ref0 != 0.0
self._has_output_adder |= ref0 != 0.0
else:
self._has_output_scaling |= np.any(ref0)
self._has_output_adder |= np.any(ref0)
if res_ref is not None:
if isscalar(res_ref):
self._has_resid_scaling |= res_ref != 1.0
else:
self._has_resid_scaling |= np.any(res_ref != 1.0)
# until we get rid of component level distributed option, handle the case where
# component distributed has been set to True but variable distributed has been set
# to False by the caller.
if distributed is not False:
if distributed is None:
distributed = False
# using ._dict below to avoid tons of deprecation warnings
distributed = distributed or ('distributed' in self.options and
self.options._dict['distributed']['val'])
if copy_shape and compute_shape:
raise ValueError(f"{self.msginfo}: Only one of 'copy_shape' or 'compute_shape' can "
"be specified.")
if copy_shape and not isinstance(copy_shape, str):
raise TypeError(f"{self.msginfo}: The copy_shape argument should be a str or None but "
f"a '{type(copy_shape).__name__}' was given.")
if compute_shape and not callable(compute_shape):
raise TypeError(f"{self.msginfo}: The compute_shape argument should be callable but "
f"a '{type(compute_shape).__name__}' was given.")
if compute_shape is not None and is_lambda(compute_shape):
compute_shape = LambdaPickleWrapper(compute_shape)
if primal_name is not None:
self._valid_name_map[name] = primal_name
metadata = {
'val': val,
'shape': shape,
'size': shape_to_len(shape),
'units': units,
'res_units': res_units,
'desc': desc,
'distributed': distributed,
'tags': make_set(tags),
'ref': format_as_float_or_array('ref', ref, flatten=True),
'ref0': format_as_float_or_array('ref0', ref0, flatten=True),
'res_ref': format_as_float_or_array('res_ref', res_ref, flatten=True, val_if_none=None),
'lower': lower,
'upper': upper,
'shape_by_conn': shape_by_conn,
'compute_shape': compute_shape,
'copy_shape': copy_shape,
}
# this will get reset later if comm size is 1
self._has_distrib_vars |= metadata['distributed']
self._has_distrib_outputs |= metadata['distributed']
# We may not know the pathname yet, so we have to use name for now, instead of abs_name.
if self._static_mode:
var_rel2meta = self._static_var_rel2meta
var_rel_names = self._static_var_rel_names
else:
var_rel2meta = self._var_rel2meta
var_rel_names = self._var_rel_names
# Disallow dupes
if name in var_rel2meta:
raise ValueError("{}: Variable name '{}' already exists.".format(self.msginfo, name))
var_rel2meta[name] = metadata
var_rel_names['output'].append(name)
self._var_added(name)
return metadata
[docs] def add_discrete_output(self, name, val, desc='', tags=None, primal_name=None):
"""
Add an output variable to the component.
Parameters
----------
name : str
Name of the variable in this component's namespace.
val : a picklable object
The initial value of the variable being added.
desc : str
Description of the variable.
tags : str or list of strs or set of strs
User defined tags that can be used to filter what gets listed when calling
list_inputs and list_outputs.
primal_name : str or None
Valid python name to represent the variable in compute_primal if 'name' is not a valid
python name.
Returns
-------
dict
Metadata for added variable.
"""
if not isinstance(name, str):
raise TypeError('%s: The name argument should be a string.' % self.msginfo)
if not _valid_var_name(name):
raise NameError("%s: '%s' is not a valid output name." % (self.msginfo, name))
if tags is not None and not isinstance(tags, (str, set, list)):
raise TypeError('%s: The tags argument should be a str, set, or list' % self.msginfo)
if primal_name is not None:
self._valid_name_map[name] = primal_name
metadata = {}
metadata.update({
'val': val,
'type': type(val),
'desc': desc,
'tags': make_set(tags)
})
if metadata['type'] == np.ndarray:
metadata.update({'shape': val.shape})
if self._static_mode:
var_rel2meta = self._static_var_rel2meta
else:
var_rel2meta = self._var_rel2meta
# Disallow dupes
if name in var_rel2meta:
raise ValueError("{}: Variable name '{}' already exists.".format(self.msginfo, name))
var_rel2meta[name] = self._var_discrete['output'][name] = metadata
self._var_added(name)
return metadata
def _var_added(self, name):
"""
Notify config that a variable has been added to this Component.
Parameters
----------
name : str
Name of the added variable.
"""
if self._problem_meta is not None and self._problem_meta['config_info'] is not None:
self._problem_meta['config_info']._var_added(self.pathname, name)
def _update_dist_src_indices(self, abs_in2out, all_abs2meta, all_abs2idx, all_sizes):
"""
Set default src_indices for any distributed inputs where they aren't set.
Parameters
----------
abs_in2out : dict
Mapping of connected inputs to their source. Names are absolute.
all_abs2meta : dict
Mapping of absolute names to metadata for all variables in the model.
all_abs2idx : dict
Dictionary mapping an absolute name to its allprocs variable index.
all_sizes : dict
Mapping of types to sizes of each variable in all procs.
Returns
-------
list
Names of inputs where src_indices were added.
"""
iproc = self.comm.rank
abs2meta_in = self._var_abs2meta['input']
all_abs2meta_in = all_abs2meta['input']
all_abs2meta_out = all_abs2meta['output']
abs_in2prom_info = self._problem_meta['abs_in2prom_info']
sizes_in = self._var_sizes['input']
sizes_out = all_sizes['output']
added_src_inds = []
# loop over continuous inputs
for i, (iname, meta_in) in enumerate(abs2meta_in.items()):
if meta_in['src_indices'] is None and iname not in abs_in2prom_info:
src = abs_in2out[iname]
dist_in = meta_in['distributed']
dist_out = all_abs2meta_out[src]['distributed']
if dist_in or dist_out:
gsize_out = all_abs2meta_out[src]['global_size']
gsize_in = all_abs2meta_in[iname]['global_size']
vout_sizes = sizes_out[:, all_abs2idx[src]]
offset = None
if gsize_out == gsize_in or (not dist_out and np.sum(vout_sizes)
== gsize_in):
# This assumes one of:
# 1) a distributed output with total size matching the total size of a
# distributed input
# 2) a non-distributed output with local size matching the total size of a
# distributed input
# 3) a non-distributed output with total size matching the total size of a
# distributed input
if dist_in:
offset = np.sum(sizes_in[:iproc, i])
end = offset + sizes_in[iproc, i]
# total sizes differ and output is distributed, so can't determine mapping
if offset is None:
self._collect_error(f"{self.msginfo}: Can't determine src_indices "
f"automatically for input '{iname}'. They must be "
"supplied manually.", ident=(self.pathname, iname))
continue
if dist_in and not dist_out:
src_shape = self._get_full_dist_shape(src, all_abs2meta_out[src]['shape'])
else:
src_shape = all_abs2meta_out[src]['global_shape']
if offset == end:
idx = np.zeros(0, dtype=INT_DTYPE)
else:
idx = slice(offset, end)
meta_in['src_indices'] = indexer(idx, flat_src=True, src_shape=src_shape)
meta_in['flat_src_indices'] = True
added_src_inds.append(iname)
return added_src_inds
def _approx_partials(self, of, wrt, method='fd', **kwargs):
"""
Inform the framework that the specified derivatives are to be approximated.
Parameters
----------
of : str or list of str
The name of the residual(s) that derivatives are being computed for.
May also contain a glob pattern.
wrt : str or list of str
The name of the variables that derivatives are taken with respect to.
This can contain the name of any input or output variable.
May also contain a glob pattern.
method : str
The type of approximation that should be used. Valid options include:
- 'fd': Finite Difference
**kwargs : dict
Keyword arguments for controlling the behavior of the approximation.
"""
self._has_approx = True
info = self._subjacs_info
for abs_key in self._matching_key_iter(of, wrt):
meta = info[abs_key]
meta['method'] = method
meta.update(kwargs)
[docs] def declare_partials(self, of, wrt, dependent=True, rows=None, cols=None, val=None,
method='exact', step=None, form=None, step_calc=None, minimum_step=None):
"""
Declare information about this component's subjacobians.
Parameters
----------
of : str or iter of str
The name of the residual(s) that derivatives are being computed for.
May also contain a glob pattern.
wrt : str or iter of str
The name of the variables that derivatives are taken with respect to.
This can contain the name of any input or output variable.
May also contain a glob pattern.
dependent : bool(True)
If False, specifies no dependence between the output(s) and the
input(s). This is only necessary in the case of a sparse global
jacobian, because if 'dependent=False' is not specified and
declare_partials is not called for a given pair, then a dense
matrix of zeros will be allocated in the sparse global jacobian
for that pair. In the case of a dense global jacobian it doesn't
matter because the space for a dense subjac will always be
allocated for every pair.
rows : ndarray of int or None
Row indices for each nonzero entry. For sparse subjacobians only.
cols : ndarray of int or None
Column indices for each nonzero entry. For sparse subjacobians only.
val : float or ndarray of float or scipy.sparse
Value of subjacobian. If rows and cols are not None, this will
contain the values found at each (row, col) location in the subjac.
method : str
The type of approximation that should be used. Valid options include:
'fd': Finite Difference, 'cs': Complex Step, 'exact': use the component
defined analytic derivatives. Default is 'exact'.
step : float
Step size for approximation. Defaults to None, in which case the approximation
method provides its default value.
form : str
Form for finite difference, can be 'forward', 'backward', or 'central'. Defaults
to None, in which case the approximation method provides its default value.
step_calc : str
Step type for computing the size of the finite difference step. It can be 'abs' for
absolute, 'rel_avg' for a size relative to the absolute value of the vector input, or
'rel_element' for a size relative to each value in the vector input. In addition, it
can be 'rel_legacy' for a size relative to the norm of the vector. For backwards
compatibilty, it can be 'rel', which is now equivalent to 'rel_avg'. Defaults to None,
in which case the approximation method provides its default value.
minimum_step : float
Minimum step size allowed when using one of the relative step_calc options.
Returns
-------
dict
Metadata dict for the specified partial(s).
"""
try:
method_func = _supported_methods[method]
except KeyError:
msg = '{}: d({})/d({}): method "{}" is not supported, method must be one of {}'
raise ValueError(msg.format(self.msginfo, of, wrt, method, sorted(_supported_methods)))
if not isinstance(of, (str, Iterable)):
raise ValueError(f"{self.msginfo}: in declare_partials, the 'of' arg must be a string "
f"or an iter of strings, but got {of}.")
if not isinstance(wrt, (str, Iterable)):
raise ValueError(f"{self.msginfo}: in declare_partials, the 'wrt' arg must be a "
f"string or an iter of strings, but got {wrt}.")
of = of if isinstance(of, str) else tuple(of)
wrt = wrt if isinstance(wrt, str) else tuple(wrt)
key = (of, wrt)
if key not in self._declared_partials_patterns:
self._declared_partials_patterns[key] = {}
meta = self._declared_partials_patterns[key]
meta['dependent'] = dependent
# If only one of rows/cols is specified
if (rows is None) ^ (cols is None):
raise ValueError('{}: d({})/d({}): If one of rows/cols is specified, then '
'both must be specified.'.format(self.msginfo, of, wrt))
if dependent:
meta['val'] = val
_val = val.data if issparse(val) else val
if np.all(_val == 0):
warn_deprecation(f'{self.msginfo}: d({of})/d({wrt}): Partial was declared to be '
f'exactly zero. This is inefficient and the declaration should '
f'be removed. In a future version of OpenMDAO this behavior '
f'will raise an error.')
if rows is not None:
rows = np.asarray(rows, dtype=INT_DTYPE)
cols = np.asarray(cols, dtype=INT_DTYPE)
# Check the length of rows and cols to catch this easy mistake and give a
# clear message.
if len(cols) != len(rows):
raise RuntimeError("{}: d({})/d({}): declare_partials has been called "
"with rows and cols, which should be arrays of equal length,"
" but rows is length {} while cols is length "
"{}.".format(self.msginfo, of, wrt, len(rows), len(cols)))
if rows.size > 0 and rows.min() < 0:
msg = '{}: d({})/d({}): row indices must be non-negative'
raise ValueError(msg.format(self.msginfo, of, wrt))
if cols.size > 0 and cols.min() < 0:
msg = '{}: d({})/d({}): col indices must be non-negative'
raise ValueError(msg.format(self.msginfo, of, wrt))
meta['rows'] = rows
meta['cols'] = cols
# Check for repeated rows/cols indices.
size = len(rows)
if size > 0:
coo = coo_matrix((np.ones(size, dtype=np.short), (rows, cols)))
dsize = coo.data.size
csc = coo.tocsc()
# csc adds values at duplicate indices together, so result will be that data
# size is less if there are duplicates
if csc.data.size < dsize:
coo = csc.tocoo()
del csc
inds = np.where(coo.data > 1.)
dups = list(zip(coo.row[inds], coo.col[inds]))
raise RuntimeError("{}: d({})/d({}): declare_partials has been called "
"with rows and cols that specify the following duplicate"
" subjacobian entries: {}.".format(self.msginfo, of, wrt,
sorted(dups)))
if method_func is not None:
# we're doing approximations
self._has_approx = True
meta['method'] = method
self._get_approx_scheme(method)
default_opts = method_func.DEFAULT_OPTIONS
else:
default_opts = ()
if step:
if 'step' in default_opts:
meta['step'] = step
else:
raise RuntimeError("{}: d({})/d({}): 'step' is not a valid option for "
"'{}'".format(self.msginfo, of, wrt, method))
if minimum_step is not None:
if 'minimum_step' in default_opts:
meta['minimum_step'] = minimum_step
else:
raise RuntimeError("{}: d({})/d({}): 'minimum_step' is not a valid option for "
"'{}'".format(self.msginfo, of, wrt, method))
if form:
if 'form' in default_opts:
meta['form'] = form
else:
raise RuntimeError("{}: d({})/d({}): 'form' is not a valid option for "
"'{}'".format(self.msginfo, of, wrt, method))
if step_calc:
if 'step_calc' in default_opts:
meta['step_calc'] = step_calc
else:
raise RuntimeError("{}: d({})/d({}): 'step_calc' is not a valid option "
"for '{}'".format(self.msginfo, of, wrt, method))
return meta
[docs] def declare_coloring(self,
wrt=_DEFAULT_COLORING_META['wrt_patterns'],
method=_DEFAULT_COLORING_META['method'],
form=None,
step=None,
per_instance=_DEFAULT_COLORING_META['per_instance'],
num_full_jacs=_DEFAULT_COLORING_META['num_full_jacs'],
tol=_DEFAULT_COLORING_META['tol'],
orders=_DEFAULT_COLORING_META['orders'],
perturb_size=_DEFAULT_COLORING_META['perturb_size'],
min_improve_pct=_DEFAULT_COLORING_META['min_improve_pct'],
show_summary=_DEFAULT_COLORING_META['show_summary'],
show_sparsity=_DEFAULT_COLORING_META['show_sparsity']):
"""
Set options for deriv coloring of a set of wrt vars matching the given pattern(s).
Parameters
----------
wrt : str or list of str
The name or names of the variables that derivatives are taken with respect to.
This can contain input names, output names, or glob patterns.
method : str
Method used to compute derivative: "fd" for finite difference, "cs" for complex step.
form : str
Finite difference form, can be "forward", "central", or "backward". Leave
undeclared to keep unchanged from previous or default value.
step : float
Step size for finite difference. Leave undeclared to keep unchanged from previous
or default value.
per_instance : bool
If True, a separate coloring will be generated for each instance of a given class.
Otherwise, only one coloring for a given class will be generated and all instances
of that class will use it.
num_full_jacs : int
Number of times to repeat partial jacobian computation when computing sparsity.
tol : float
Tolerance used to determine if an array entry is nonzero during sparsity determination.
orders : int
Number of orders above and below the tolerance to check during the tolerance sweep.
perturb_size : float
Size of input/output perturbation during generation of sparsity.
min_improve_pct : float
If coloring does not improve (decrease) the number of solves more than the given
percentage, coloring will not be used.
show_summary : bool
If True, display summary information after generating coloring.
show_sparsity : bool
If True, display sparsity with coloring info after generating coloring.
"""
super().declare_coloring(wrt, method, form, step, per_instance,
num_full_jacs,
tol, orders, perturb_size, min_improve_pct,
show_summary, show_sparsity)
# create approx partials for all matches
meta = self.declare_partials('*', wrt, method=method, step=step, form=form)
meta['coloring'] = True
[docs] def set_check_partial_options(self, wrt, method='fd', form=None, step=None, step_calc=None,
minimum_step=None, directional=False):
"""
Set options that will be used for checking partial derivatives.
Parameters
----------
wrt : str or list of str
The name or names of the variables that derivatives are taken with respect to.
This can contain the name of any input or output variable.
May also contain a glob pattern.
method : str
Method for check: "fd" for finite difference, "cs" for complex step.
form : str
Finite difference form for check, can be "forward", "central", or "backward". Leave
undeclared to keep unchanged from previous or default value.
step : float
Step size for finite difference check. Leave undeclared to keep unchanged from previous
or default value.
step_calc : str
Step type for computing the size of the finite difference step. It can be 'abs' for
absolute, 'rel_avg' for a size relative to the absolute value of the vector input, or
'rel_element' for a size relative to each value in the vector input. In addition, it
can be 'rel_legacy' for a size relative to the norm of the vector. For backwards
compatibilty, it can be 'rel', which is now equivalent to 'rel_avg'. Defaults to None,
in which case the approximation method provides its default value..
minimum_step : float
Minimum step size allowed when using one of the relative step_calc options.
directional : bool
Set to True to perform a single directional derivative for each vector variable in the
pattern named in wrt.
"""
supported_methods = ('fd', 'cs')
if method not in supported_methods:
msg = "{}: Method '{}' is not supported, method must be one of {}"
raise ValueError(msg.format(self.msginfo, method, supported_methods))
if step and not isinstance(step, (int, float)):
msg = "{}: The value of 'step' must be numeric, but '{}' was specified."
raise ValueError(msg.format(self.msginfo, step))
supported_step_calc = ('abs', 'rel', 'rel_legacy', 'rel_avg', 'rel_element')
if step_calc and step_calc not in supported_step_calc:
msg = "{}: The value of 'step_calc' must be one of {}, but '{}' was specified."
raise ValueError(msg.format(self.msginfo, supported_step_calc, step_calc))
if not isinstance(wrt, (str, list, tuple)):
msg = "{}: The value of 'wrt' must be a string or list of strings, but a type " \
"of '{}' was provided."
raise ValueError(msg.format(self.msginfo, type(wrt).__name__))
if not isinstance(directional, bool):
msg = "{}: The value of 'directional' must be True or False, but a type " \
"of '{}' was provided."
raise ValueError(msg.format(self.msginfo, type(directional).__name__))
wrt_list = [wrt] if isinstance(wrt, str) else wrt
self._declared_partial_checks.append((wrt_list, method, form, step, step_calc,
minimum_step, directional))
def _get_check_partial_options(self):
"""
Return dictionary of partial options with pattern matches processed.
This is called by check_partials.
Returns
-------
dict(wrt: (options))
Dictionary keyed by name with tuples of options (method, form, step, step_calc,
minimum_step, directional)
"""
if not self._declared_partial_checks:
return {}
opts = {}
wrt = self._get_partials_wrts()
invalid_wrt = []
matrix_free = self.matrix_free
if matrix_free:
n_directional = 0
for data_tup in self._declared_partial_checks:
wrt_list, method, form, step, step_calc, minimum_step, directional = data_tup
for pattern in wrt_list:
matches = find_matches(pattern, wrt)
# if a non-wildcard var name was specified and not found, save for later Exception
if len(matches) == 0 and _valid_var_name(pattern):
invalid_wrt.append(pattern)
for match in matches:
if match in opts:
opt = opts[match]
# New assignments take precedence
keynames = ['method', 'form', 'step', 'step_calc', 'minimum_step',
'directional']
for name, value in zip(keynames,
[method, form, step, step_calc, minimum_step,
directional]):
if value is not None:
opt[name] = value
else:
opts[match] = {'method': method,
'form': form,
'step': step,
'step_calc': step_calc,
'minimum_step': minimum_step,
'directional': directional}
if matrix_free and directional:
n_directional += 1
if invalid_wrt:
msg = "{}: Invalid 'wrt' variables specified for check_partial options: {}."
raise ValueError(msg.format(self.msginfo, invalid_wrt))
if matrix_free:
if n_directional > 0 and n_directional < len(wrt):
msg = "{}: For matrix free components, directional should be set to True for " + \
"all inputs."
raise ValueError(msg.format(self.msginfo))
return opts
def _get_approx_partial_options(self, key, method='fd', checkopts=None):
# TODO: extend this to Groups
approx = self._get_approx_scheme(method)
options = approx.DEFAULT_OPTIONS.copy()
if checkopts is None:
options.update(approx._wrt_meta)
else:
options.update(checkopts)
options.update(approx._wrt_meta)
abs_key = rel_key2abs_key(self, key)
if abs_key in self._subjacs_info:
meta = self._subjacs_info[abs_key]
options.update({k: v for k, v in meta.items() if v is not None and k in options})
return options
def _resolve_partials_patterns(self, of, wrt, pattern_meta):
"""
Store subjacobian metadata for specific of, wrt pairs after resolving glob patterns.
Parameters
----------
of : tuple of str
The names of the residuals that derivatives are being computed for.
May also contain glob patterns.
wrt : tuple of str
The names of the variables that derivatives are taken with respect to.
This can contain the name of any input or output variable.
May also contain glob patterns.
pattern_meta : dict
Metadata dict specifying shape, and/or approx properties, keyed by (of, wrt) as
described above.
"""
val = pattern_meta['val'] if 'val' in pattern_meta else None
is_scalar = isscalar(val)
dependent = pattern_meta['dependent']
matfree = self.matrix_free
if dependent:
if 'rows' in pattern_meta and pattern_meta['rows'] is not None: # sparse list format
rows = pattern_meta['rows']
cols = pattern_meta['cols']
if is_scalar:
val = np.full(rows.size, val, dtype=float)
is_scalar = False
elif val is not None:
# np.promote_types will choose the smallest dtype that can contain
# both arguments
val = atleast_1d(val)
safe_dtype = promote_types(val.dtype, float)
val = val.astype(safe_dtype, copy=False)
if rows.shape != val.shape:
raise ValueError('{}: d({})/d({}): If rows and cols are specified, val '
'must be a scalar or have the same shape, val: {}, '
'rows/cols: {}'.format(self.msginfo, of, wrt,
val.shape, rows.shape))
elif not matfree:
val = np.zeros_like(rows, dtype=float)
if rows.size > 0:
rows_max = rows.max()
cols_max = cols.max()
else:
rows_max = cols_max = 0
else:
if val is not None and not is_scalar and not issparse(val):
val = atleast_2d(val)
val = val.astype(promote_types(val.dtype, float), copy=False)
rows_max = cols_max = 0
rows = None
cols = None
abs2meta_in = self._var_abs2meta['input']
abs2meta_out = self._var_abs2meta['output']
is_array = isinstance(val, ndarray)
patmeta = dict(pattern_meta)
patmeta_not_none = {k: v for k, v in pattern_meta.items() if v is not None}
for abs_key in self._matching_key_iter(of, '*' if wrt is None else wrt):
if not dependent:
if abs_key in self._subjacs_info:
del self._subjacs_info[abs_key]
continue
if abs_key in self._subjacs_info:
meta = self._subjacs_info[abs_key]
meta.update(patmeta_not_none)
if 'rows' in meta and meta['rows'] is not None:
rows = meta['rows']
if 'cols' in meta and meta['cols'] is not None:
cols = meta['cols']
else:
meta = patmeta.copy()
of, wrt = abs_key
meta['rows'] = rows
meta['cols'] = cols
csz = abs2meta_in[wrt]['size'] if wrt in abs2meta_in else abs2meta_out[wrt]['size']
meta['shape'] = shape = (abs2meta_out[of]['size'], csz)
dist_out = abs2meta_out[of]['distributed']
if wrt in abs2meta_in:
dist_in = abs2meta_in[wrt]['distributed']
else:
dist_in = abs2meta_out[wrt]['distributed']
if dist_in and not dist_out and not matfree:
rel_key = abs_key2rel_key(self, abs_key)
raise RuntimeError(f"{self.msginfo}: component has defined partial {rel_key} "
"which is a non-distributed output wrt a distributed input."
" This is only supported using the matrix free API.")
if shape[0] == 0 or shape[1] == 0:
msg = "{}: '{}' is an array of size 0"
if shape[0] == 0:
if dist_out:
# distributed vars are allowed to have zero size inputs on some procs
rows_max = -1
else:
# non-distributed vars are not allowed to have zero size inputs
raise ValueError(msg.format(self.msginfo, of))
if shape[1] == 0:
if not dist_in:
# non-distributed vars are not allowed to have zero size outputs
raise ValueError(msg.format(self.msginfo, wrt))
else:
# distributed vars are allowed to have zero size outputs on some procs
cols_max = -1
if val is None and not matfree:
# we can only get here if rows is None (we're not sparse list format)
meta['val'] = np.zeros(shape)
elif is_array:
if rows is None and val.shape != shape and val.size == shape[0] * shape[1]:
meta['val'] = val = val.copy().reshape(shape)
else:
meta['val'] = val.copy()
elif is_scalar:
meta['val'] = np.full(shape, val, dtype=float)
else:
meta['val'] = val
if rows_max >= shape[0] or cols_max >= shape[1]:
of, wrt = abs_key2rel_key(self, abs_key)
raise ValueError(f"{self.msginfo}: d({of})/d({wrt}): Expected {shape[0]}x"
f"{shape[1]} but declared at least {rows_max + 1}x"
f"{cols_max + 1}")
self._check_partials_meta(abs_key, meta['val'],
shape if rows is None else (rows.shape[0], 1))
self._subjacs_info[abs_key] = meta
def _column_iotypes(self):
"""
Return a tuple of the iotypes that make up columns of the jacobian.
Returns
-------
tuple of the form ('input',)
The iotypes that make up columns of the jacobian.
"""
return ('input',)
def _get_partials_wrts(self):
"""
Get list of 'wrt' variables that form the partial jacobian.
Returns
-------
list
List of 'wrt' relative variable names.
"""
namelists = [self._var_rel_names[io] for io in self._column_iotypes()]
return [n for n in chain(*namelists)]
def _get_partials_ofs(self, use_resname=False):
"""
Get lists of 'of' variables that form the partial jacobian.
Parameters
----------
use_resname : bool
Ignored.
Returns
-------
list
List of 'of' relative variable names.
"""
return list(self._var_rel_names['output'])
def _matching_key_iter(self, of_patterns, wrt_patterns, use_resname=False):
"""
Iterate over all combinations of matching keys for the given patterns.
Parameters
----------
of_patterns : list of str
List of variable names and/or glob patterns for the 'of' variables.
wrt_patterns : list of str
List of variable names and/or glob patterns for the 'wrt' variables.
use_resname : bool, optional
If True, match of_patterns against residuals instead of outputs.
Yields
------
tuple
A tuple of matching keys, where the first element is the 'of' key and the second
element is the 'wrt' key. Both are absolute names.
"""
of_bundles = self._find_of_matches(of_patterns, use_resname=use_resname)
wrt_bundles = self._find_wrt_matches(wrt_patterns)
for of_bundle, wrt_bundle in product(of_bundles, wrt_bundles):
of_pattern, of_matches = of_bundle
wrt_pattern, wrt_matches = wrt_bundle
if not of_matches:
raise ValueError(f'{self.msginfo}: No matches were found for of="{of_pattern}"')
if not wrt_matches:
raise ValueError(f'{self.msginfo}: No matches were found for wrt="{wrt_pattern}"')
yield from abs_key_iter(self, of_matches, wrt_matches)
def _find_of_matches(self, pattern, use_resname=False):
"""
Find all matches for the given 'of' pattern.
Parameters
----------
pattern : str
Glob pattern or relative variable name.
use_resname : bool
If True, match residual names instead of output names.
Returns
-------
list
List of tuples of the form (abs_name, meta) where abs_name is the absolute name of the
matching variable and meta is the metadata for that variable.
"""
of_list = [pattern] if isinstance(pattern, str) else pattern
return [(pattern, find_matches(pattern, self._get_partials_ofs(use_resname=use_resname)))
for pattern in of_list]
def _find_wrt_matches(self, pattern):
"""
Find all matches for the given 'wrt' pattern.
Parameters
----------
pattern : str
Glob pattern or relative variable name.
Returns
-------
list
List of tuples of the form (abs_name, meta) where abs_name is the absolute name of the
matching variable and meta is the metadata for that variable.
"""
patterns = [pattern] if isinstance(pattern, str) else pattern
return [(pattern, find_matches(pattern, self._get_partials_wrts())) for pattern in patterns]
def _check_partials_meta(self, abs_key, val, shape):
"""
Check a given partial derivative and metadata for the correct shapes.
Parameters
----------
abs_key : tuple(str, str)
The of/wrt pair (given absolute names) defining the partial derivative.
val : ndarray
Subjac value.
shape : tuple
Expected shape of val.
"""
out_size, in_size = shape
if in_size == 0 and self.comm.rank != 0: # 'inactive' component
return
if val is not None:
val_shape = val.shape
if len(val_shape) == 1:
val_out, val_in = val_shape[0], 1
else:
val_out, val_in = val.shape
if val_out > out_size or val_in > in_size:
of, wrt = abs_key2rel_key(self, abs_key)
msg = '{}: d({})/d({}): Expected {}x{} but val is {}x{}'
raise ValueError(msg.format(self.msginfo, of, wrt, out_size, in_size,
val_out, val_in))
def _set_approx_partials_meta(self):
"""
Add approximations for those partials registered with method=fd or method=cs.
"""
self._get_static_wrt_matches()
subjacs = self._subjacs_info
wrtset = set()
subjac_keys = self._get_approx_subjac_keys()
# go through subjac keys in reverse and only add approx for the last of each wrt
# (this prevents warnings that could confuse users)
for i in range(len(subjac_keys) - 1, -1, -1):
key = subjac_keys[i]
if key[1] not in wrtset:
wrtset.add(key[1])
meta = subjacs[key]
self._approx_schemes[meta['method']].add_approximation(key, self, meta)
def _guess_nonlinear(self):
"""
Provide initial guess for states.
Does nothing on any non-implicit component.
"""
pass
def _clear_iprint(self):
"""
Clear out the iprint stack from the solvers.
Components don't have nested solvers, so do nothing to prevent errors.
"""
pass
def _check_first_linearize(self):
if self._first_call_to_linearize:
self._first_call_to_linearize = False # only do this once
if coloring_mod._use_partial_sparsity:
self._get_coloring()
if self._jacobian is not None:
self._jacobian._restore_approx_sparsity()
def _resolve_src_inds(self):
abs2prom = self._resolver.abs2prom
abs_in2prom_info = self._problem_meta['abs_in2prom_info']
all_abs2meta_in = self._var_allprocs_abs2meta['input']
abs2meta_in = self._var_abs2meta['input']
conns = self._problem_meta['model_ref']()._conn_global_abs_in2out
all_abs2meta_out = self._problem_meta['model_ref']()._var_allprocs_abs2meta['output']
for tgt, meta in abs2meta_in.items():
if tgt in abs_in2prom_info:
pinfo = abs_in2prom_info[tgt][-1] # component always last in the plist
if pinfo is not None:
inds, flat, shape = pinfo
if inds is not None:
all_abs2meta_in[tgt]['has_src_indices'] = True
meta['src_shape'] = shape = all_abs2meta_out[conns[tgt]]['global_shape']
if inds._flat_src:
meta['flat_src_indices'] = True
elif meta['flat_src_indices'] is None:
meta['flat_src_indices'] = flat
try:
if not isinstance(inds, Indexer):
meta['src_indices'] = inds = indexer(inds, flat_src=flat,
src_shape=shape)
else:
meta['src_indices'] = inds = inds.copy()
inds.set_src_shape(shape)
self._var_prom2inds[abs2prom(tgt, iotype='input')] = \
[shape, inds, flat]
except Exception:
type_exc, exc, tb = sys.exc_info()
self._collect_error(f"When accessing '{conns[tgt]}' with src_shape "
f"{shape} from '{pinfo.prom_path()}' using "
f"src_indices {inds}: {exc}", exc_type=type_exc,
tback=tb, ident=(conns[tgt], tgt))
def _check_consistent_serial_dinputs(self, nz_dist_outputs):
"""
Check consistency across ranks for serial d_inputs variables.
This is used primarily to test that `compute_jacvec_product` and `apply_linear` methods
follow the OpenMDAO convention that in reverse mode, the component should perform
'allreduce' sorts of operations only for derivatives of distributed outputs with-respect-to
serial inputs. This should result in serial input derivatives being consistent across all
ranks in the Component's communicator.
Parameters
----------
nz_dist_outputs : set or list
Set of distributed outputs with nonzero values for the most recent _apply_linear call.
"""
if not self.checking or not self._has_distrib_outputs or self.comm.size < 2:
return
if self._serial_idxs is None:
ranges = defaultdict(list)
output_len = 0 if self.is_explicit() else len(self._outputs)
for _, offset, end, vec, slc, dist_sizes in self._jac_wrt_iter():
if dist_sizes is None: # not distributed
if offset != end:
if vec is self._outputs:
ranges[vec].append(range(offset, end))
else:
ranges[vec].append(range(offset - output_len, end - output_len))
self._serial_idxs = []
for vec, rlist in ranges.items():
if rlist:
self._serial_idxs.append((vec, np.hstack(rlist)))
for vec, inds in self._serial_idxs:
# _jac_wrt_iter gives us _input and possibly _output vecs (for implicit comps), but we
# want to check _dinputs and _doutputs
v = self._dinputs if vec is self._inputs else self._doutputs
result = inconsistent_across_procs(self.comm, v.asarray()[inds])
if self.comm.rank == 0 and np.any(result):
bad_inds = np.arange(len(v), dtype=INT_DTYPE)[inds][result]
bad_mask = np.zeros(len(v), dtype=bool)
bad_mask[bad_inds] = True
for inname, start, stop in v.ranges():
if np.any(bad_mask[start:stop]):
for outname in nz_dist_outputs:
key = (outname, inname)
self._inconsistent_keys.add(key)
def _get_dist_nz_dresids(self):
"""
Get names of distributed resids that are non-zero prior to computing derivatives.
This should only be called when 'rev' mode is active.
Returns
-------
list of str
List of names of distributed resids that have nonzero entries.
"""
nzresids = []
dresids = self._dresiduals.asarray()
for of, start, end, _, dist_sizes in self._jac_of_iter():
if dist_sizes is not None:
if np.any(dresids[start:end]):
nzresids.append(of)
full_nzresids = set()
if self.comm.rank == 0:
for nzoutlist in self.comm.gather(nzresids, root=0):
full_nzresids.update(nzoutlist)
return full_nzresids
self.comm.gather(nzresids, root=0)
return nzresids
def _get_graph_node_meta(self):
"""
Return metadata to add to this system's graph node.
Returns
-------
dict
Metadata for this system's graph node.
"""
meta = super()._get_graph_node_meta()
meta['base'] = 'ExplicitComponent' if self.is_explicit() else 'ImplicitComponent'
return meta
[docs] def compute_fd_jac(self, jac, method='fd'):
"""
Force the use of finite difference to compute a jacobian.
This can be used to compute sparsity for a component that computes derivatives analytically
in order to check the accuracy of the declared sparsity.
Parameters
----------
jac : Jacobian
The Jacobian object that will contain the computed jacobian.
method : str
The type of finite difference to perform. Valid options are 'fd' for forward difference,
or 'cs' for complex step.
"""
if method == 'jax':
method = 'fd'
fd_methods = {'fd': _supported_methods['fd'], 'cs': _supported_methods['cs']}
try:
approximation = fd_methods[method]()
except KeyError:
raise ValueError(f"Method '{method}' is not a recognized finite difference method.")
# these are relative names
ofs = self._get_partials_ofs()
wrts = self._get_partials_wrts()
local_opts = self._get_check_partial_options()
added_wrts = set()
for rel_key in product(ofs, wrts):
fd_options = self._get_approx_partial_options(rel_key, method=method,
checkopts=local_opts)
abs_key = rel_key2abs_key(self, rel_key)
# prevent adding multiple approxs with same wrt (and confusing users with warnings)
if abs_key[1] not in added_wrts:
approximation.add_approximation(abs_key, self, fd_options)
added_wrts.add(abs_key[1])
# Perform the FD here.
with self._unscaled_context(outputs=[self._outputs], residuals=[self._residuals]):
approximation.compute_approximations(self, jac=jac)
[docs] def check_sparsity(self, method='fd', max_nz=90., out_stream=_DEFAULT_OUT_STREAM):
"""
Check the sparsity of the computed jacobian against the declared sparsity.
Check is skipped if one of the dimensions of the jacobian is 1 or if the percentage of
nonzeros in the computed jacobian is greater than max_nz%.
Parameters
----------
method : str
The type of finite difference to perform. Valid options are 'fd' for forward difference,
or 'cs' for complex step.
max_nz : float
If the percentage of nonzeros in a sub-jacobian exceeds this, no warning is issued if
the computed sparsity does not match the declared sparsity.
out_stream : file-like object
Where to send the output. If None, output will be suppressed.
Returns
-------
list
A list of tuples, one for each subjacobian that has a mismatch between the computed
sparsity and the declared sparsity. Each tuple has the form (of, wrt, computed_rows,
computed_cols, declared_rows, declared_cols, shape, pct_nonzero).
"""
if out_stream == _DEFAULT_OUT_STREAM:
out_stream = sys.stdout
def rowsizeiter():
for of, start, end, _, _ in self._jac_of_iter():
yield of, end - start
def colsizeiter():
for wrt, start, end, _, _, _ in self._jac_wrt_iter():
yield wrt, end - start
sparsity, _ = self.compute_fd_sparsity(method=method)
prefix = self.pathname + '.'
plen = len(prefix)
ret = []
for of, wrt, nzrows, nzcols, shape in submat_sparsity_iter(rowsizeiter(), colsizeiter(),
sparsity.row, sparsity.col,
sparsity.shape):
if 1 in shape:
continue
key = (of, wrt)
if key in self._subjacs_info:
meta = self._subjacs_info[key]
computed = sorted(zip(nzrows, nzcols))
if meta['rows'] is None:
rows = []
cols = []
declared = []
else:
rows = meta['rows']
cols = meta['cols']
declared = sorted(zip(rows, cols))
if declared != computed:
pct_nonzero = 100. * len(nzrows) / (shape[0] * shape[1])
if pct_nonzero > max_nz:
continue
if shape[0] > 200 or shape[1] > 200:
mstr = "Sparsity matrix too large to show."
else:
stream = StringIO()
val_map = {0: '.', 1: 'C', 3: 'D', 4: 'x'}
sparsity_diff_viz(csr_matrix((np.ones(len(nzrows)), (nzrows, nzcols)),
shape=shape, dtype=bool),
csr_matrix((np.ones(len(rows)), (rows, cols)),
shape=shape, dtype=bool),
val_map=val_map,
stream=stream)
mstr = stream.getvalue()
wrn = (f"{self.msginfo}:\n(D)eclared sparsity pattern != (c)omputed sparsity "
f"pattern for sub-jacobian ({of[plen:]}, {wrt[plen:]}) with shape "
f"{shape} and {pct_nonzero:.2f}% nonzeros:\n{mstr}\n")
ret.append((of, wrt, nzrows, nzcols, rows, cols, shape, pct_nonzero, wrn))
if out_stream is not None:
print(wrn, file=out_stream)
return ret
def _check_fds_differ(self, method, step, form, step_calc, minimum_step):
# Check to make sure the method and settings used for checking
# is different from the method used to calc the derivatives
# Could do this later in this method but at that point some computations could have been
# made and it would just waste time before the user is told there is an error and the
# program errs out
requested_method = method
alloc_complex = self._outputs._alloc_complex
local_opts = self._get_check_partial_options()
for keypats, meta in self._declared_partials_patterns.items():
# Get the complete set of options, including defaults
# for the computing of the derivs for this component
if 'method' not in meta:
meta_with_defaults = {}
meta_with_defaults['method'] = 'exact'
elif meta['method'] == 'cs':
meta_with_defaults = ComplexStep.DEFAULT_OPTIONS.copy()
else:
meta_with_defaults = FiniteDifference.DEFAULT_OPTIONS.copy()
meta_with_defaults.update(meta)
_, wrtpats = keypats
# For each of the partials, check to see if the
# check partials options are different than the options used to compute
# the partials
for _, wrtvars in self._find_wrt_matches(wrtpats):
for var in wrtvars:
# we now have individual vars like 'x'
# get the options for checking partials
fd_options, _ = _get_fd_options(var, requested_method, local_opts, step,
form, step_calc, alloc_complex,
minimum_step)
# compare the compute options to the check options
if fd_options['method'] != meta_with_defaults['method']:
all_same = False
else:
all_same = True
if fd_options['method'] == 'fd':
option_names = ['form', 'step', 'step_calc', 'minimum_step',
'directional']
else:
option_names = ['step', 'directional']
for name in option_names:
if fd_options[name] != meta_with_defaults[name]:
all_same = False
break
if all_same:
msg = (f"Checking partials with respect to variable '{var}' in component "
f"'{self.pathname}' using the same method and options as are used "
"to compute the component's derivatives will not provide any "
"relevant information on the accuracy.\n"
"To correct this, change the options to do the "
"check_partials using either:\n"
" - arguments to Problem.check_partials.\n"
" - arguments to Component.set_check_partial_options")
issue_warning(msg, prefix=self.msginfo,
category=OMInvalidCheckDerivativesOptionsWarning)
[docs] def check_partials(self, out_stream=_DEFAULT_OUT_STREAM,
compact_print=False, abs_err_tol=0.0, rel_err_tol=1e-6,
method='fd', step=None, form='forward', step_calc='abs',
minimum_step=1e-12, force_dense=True, show_only_incorrect=False,
show_worst=True):
"""
Check partial derivatives comprehensively for this component.
Parameters
----------
out_stream : file-like object
Where to send human readable output. By default it goes to stdout.
Set to None to suppress.
compact_print : bool
Set to True to just print the essentials, one line per input-output pair.
abs_err_tol : float
Threshold value for absolute error. Errors about this value will have a '*' displayed
next to them in output, making them easy to search for. Default is 1.0E-6.
rel_err_tol : float
Threshold value for relative error. Errors about this value will have a '*' displayed
next to them in output, making them easy to search for. Note at times there may be a
significant relative error due to a minor absolute error. Default is 1.0E-6.
method : str
Method, 'fd' for finite difference or 'cs' for complex step. Default is 'fd'.
step : None, float, or list/tuple of float
Step size(s) for approximation. Default is None, which means 1e-6 for 'fd' and 1e-40 for
'cs'.
form : str
Form for finite difference, can be 'forward', 'backward', or 'central'. Default
'forward'.
step_calc : str
Step type for computing the size of the finite difference step. It can be 'abs' for
absolute, 'rel_avg' for a size relative to the absolute value of the vector input, or
'rel_element' for a size relative to each value in the vector input. In addition, it
can be 'rel_legacy' for a size relative to the norm of the vector. For backwards
compatibilty, it can be 'rel', which is now equivalent to 'rel_avg'. Defaults to None,
in which case the approximation method provides its default value.
minimum_step : float
Minimum step size allowed when using one of the relative step_calc options.
force_dense : bool
If True, analytic derivatives will be coerced into arrays. Default is True.
show_only_incorrect : bool, optional
Set to True if output should print only the subjacs found to be incorrect.
show_worst : bool, optional
Set to False to suppress the display of the worst subjac.
Returns
-------
tuple of the form (derivs_dict, worst)
Where derivs_dict is a dict, where the top key is the component pathname.
Under the top key, the subkeys are the (of, wrt) keys of the subjacs.
Within the (of, wrt) entries are the following keys:
'tol violation', 'magnitude', 'J_fd', 'J_fwd', 'J_rev', 'vals_at_max_error',
and 'rank_inconsistent'.
For 'J_fd', 'J_fwd', 'J_rev' the value is a numpy array representing the computed
Jacobian for the three different methods of computation.
For 'tol violation' and 'vals_at_max_error' the value is a
tuple containing values for forward - fd, reverse - fd, forward - reverse. For
'magnitude' the value is a tuple indicating the maximum magnitude of values found in
Jfwd, Jrev, and Jfd.
The boolean 'rank_inconsistent' indicates if the derivative wrt a serial variable is
inconsistent across MPI ranks.
worst is either None or a tuple of the form (error, table_row, header)
where error is the max error found, table_row is the formatted table row
containing the max error, and header is the formatted table header. 'worst'
is not None only if compact_print is True.
"""
if out_stream == _DEFAULT_OUT_STREAM:
out_stream = sys.stdout
if self._problem_meta is None:
if self._problem_meta['setup_status'] < _SetupStatus.POST_FINAL_SETUP:
raise RuntimeError(f"{self.msginfo}: Can't check_partials before final_setup. Also,"
" make sure to set valid input values before calling "
"check_partials either manually or by running the model.")
self._check_fds_differ(method, step, form, step_calc, minimum_step)
# Make sure we're in a valid state
self.run_apply_nonlinear()
input_cache = self._inputs.asarray(copy=True)
output_cache = self._outputs.asarray(copy=True)
local_opts = self._get_check_partial_options()
if self.matrix_free:
directions = ('fwd', 'rev')
else:
# TODO: replace 'fwd' with self.best_partial_deriv_direction(). Currently fails
# when it equals 'rev' for directional derivatives.
directions = ('fwd',) # rev same as fwd for analytic jacobians
self.run_linearize(sub_do_ln=False)
nondep_derivs = set()
of_list = self._get_partials_ofs()
wrt_list = self._get_partials_wrts()
axis = {'fwd': 1, 'rev': 0}
mfree_directions = {}
abs2meta_in = self._var_allprocs_abs2meta['input']
abs2meta_out = self._var_allprocs_abs2meta['output']
partials_data = defaultdict(dict)
requested_method = method
probmeta = self._problem_meta
prefix = self.pathname + '.'
for mode in directions:
jac_key = 'J_' + mode
with self._unscaled_context():
# Matrix-free components need to calculate Jacobian by matrix-vector product.
if self.matrix_free:
dstate = self._doutputs
if mode == 'fwd':
dinputs = self._dinputs
doutputs = self._dresiduals
in_list = wrt_list
out_list = of_list
else:
dinputs = self._dresiduals
doutputs = self._dinputs
in_list = of_list
out_list = wrt_list
for inp in in_list:
inp_abs = prefix + inp
if mode == 'fwd':
directional = inp in local_opts and local_opts[inp]['directional']
else:
directional = len(mfree_directions) > 0
try:
flat_view = dinputs._abs_get_val(inp_abs)
except KeyError:
# Implicit state
flat_view = dstate._abs_get_val(inp_abs)
if directional:
n_in = 1
idxs = range(1)
if inp in mfree_directions:
perturb = mfree_directions[inp]
else:
perturb = 2.0 * np.random.random(len(flat_view)) - 1.0
mfree_directions[inp] = perturb
else:
n_in = len(flat_view)
idxs = LocalRangeIterable(self, inp_abs, use_vec_offset=False)
perturb = 1.0
for idx in idxs:
dinputs.set_val(0.0)
dstate.set_val(0.0)
if directional:
flat_view[:] = perturb
elif idx is not None:
flat_view[idx] = perturb
# Matrix Vector Product
probmeta['checking'] = True
try:
self.run_apply_linear(mode)
finally:
probmeta['checking'] = False
for out in out_list:
out_abs = prefix + out
try:
derivs = doutputs._abs_get_val(out_abs)
except KeyError:
# Implicit state
derivs = dstate._abs_get_val(out_abs)
if mode == 'fwd':
key = out, inp
deriv = partials_data[key]
# Allocate first time
if jac_key not in deriv:
shape = (len(derivs), n_in)
deriv[jac_key] = np.zeros(shape)
if idx is not None:
deriv[jac_key][:, idx] = derivs
else: # rev
key = inp, out
deriv = partials_data[key]
if directional:
# Dot product test for adjoint validity.
m = mfree_directions[out]
d = mfree_directions[inp]
mhat = derivs
dhat = deriv['J_fwd'][:, idx]
deriv['directional_fwd_rev'] = (dhat.dot(d), mhat.dot(m))
else:
meta = abs2meta_in[out_abs] if out_abs in abs2meta_in \
else abs2meta_out[out_abs]
if not meta['distributed']: # serial input or state
if inconsistent_across_procs(self.comm, derivs,
return_array=False):
deriv['rank_inconsistent'] = True
# Allocate first time
if jac_key not in deriv:
shape = (n_in, len(derivs))
deriv[jac_key] = np.zeros(shape)
if idx is not None:
deriv[jac_key][idx, :] = derivs
# This component already has a Jacobian with calculated derivatives.
else:
subjacs = self._jacobian._subjacs_info
for rel_key in product(of_list, wrt_list):
abs_key = rel_key2abs_key(self, rel_key)
of, wrt = abs_key
if mode == 'fwd':
inp = rel_key[1]
directional = inp in local_opts and local_opts[inp]['directional']
else:
directional = len(mfree_directions) > 0
if wrt in self._var_abs2meta['input']:
wrt_meta = self._var_abs2meta['input'][wrt]
else:
wrt_meta = self._var_abs2meta['output'][wrt]
copy = True
# No need to calculate partials; they are already stored
try:
deriv_value = subjacs[abs_key]['val']
rows = subjacs[abs_key]['rows']
except KeyError:
rows = None
# Missing derivatives are assumed 0.
in_size = 1 if directional else wrt_meta['size']
out_size = self._var_abs2meta['output'][of]['size']
deriv_value = None
copy = False
# Testing for pairs that are not dependent so that we suppress printing them
# unless the fd is non zero.
# Note: subjacs_info is empty for undeclared partials, which is the default
# behavior now.
try:
if not subjacs[abs_key]['dependent']:
nondep_derivs.add(rel_key)
except KeyError:
nondep_derivs.add(rel_key)
if force_dense:
if rows is not None:
in_size = wrt_meta['size']
out_size = self._var_abs2meta['output'][of]['size']
# if a scalar value is provided (in declare_partials),
# expand to the correct size array value for zipping
if deriv_value.size == 1:
deriv_value *= np.ones(rows.size)
deriv_value = \
coo_matrix((deriv_value, (rows, subjacs[abs_key]['cols'])),
shape=(out_size, in_size))
if directional:
deriv_value = \
np.atleast_2d(deriv_value.sum(axis=axis[mode]))
if deriv_value.shape[0] < deriv_value.shape[1]:
deriv_value = deriv_value.T
else:
deriv_value = deriv_value.toarray()
copy = False
elif issparse(deriv_value):
if directional:
deriv_value = \
np.atleast_2d(deriv_value.sum(axis=axis[mode])).T
if deriv_value.shape[0] < deriv_value.shape[1]:
deriv_value = deriv_value.T
else:
deriv_value = deriv_value.toarray()
copy = False
elif directional:
deriv_value = \
np.atleast_2d(np.sum(deriv_value, axis=axis[mode])).T
if copy:
deriv_value = deriv_value.copy()
if deriv_value is not None:
partials_data[rel_key][jac_key] = deriv_value
self._inputs.set_val(input_cache)
self._outputs.set_val(output_cache)
all_fd_options = {}
of = self._get_partials_ofs()
wrt = self._get_partials_wrts()
if step is None or isinstance(step, (float, int)):
steps = [step]
else:
steps = step
could_not_cs = False
actual_steps = defaultdict(list)
alloc_complex = self._outputs._alloc_complex
for step in steps:
self.run_apply_nonlinear()
approximations = {'fd': FiniteDifference(), 'cs': ComplexStep()}
added_wrts = set()
# Load up approximation objects with the requested settings.
for rel_key in product(of, wrt):
abs_key = rel_key2abs_key(self, rel_key)
local_wrt = rel_key[1]
fd_options, no_cs = _get_fd_options(local_wrt, requested_method,
local_opts, step, form, step_calc,
alloc_complex, minimum_step)
if no_cs:
could_not_cs = True
actual_steps[rel_key].append(fd_options['step'])
# Determine if fd or cs.
method = requested_method
all_fd_options[local_wrt] = fd_options
if local_wrt in mfree_directions:
vector = mfree_directions.get(local_wrt)
else:
vector = None
# prevent adding multiple approxs with same wrt (and confusing users with
# warnings)
if abs_key[1] not in added_wrts:
approximations[fd_options['method']].add_approximation(abs_key, self,
fd_options,
vector=vector)
added_wrts.add(abs_key[1])
approx_jac = _CheckingJacobian(self)
for approximation in approximations.values():
# Perform the FD here.
approximation.compute_approximations(self, jac=approx_jac)
for abs_key, partial in approx_jac.items():
rel_key = abs_key2rel_key(self, abs_key)
deriv = partials_data[rel_key]
subjacs_info = approx_jac._subjacs_info[abs_key]
_of, _wrt = rel_key
if 'J_fd' not in deriv:
deriv['J_fd'] = []
deriv['steps'] = []
deriv['J_fd'].append(partial)
deriv['steps'] = actual_steps[rel_key]
deriv['rows'] = subjacs_info['rows']
deriv['cols'] = subjacs_info['cols']
if 'uncovered_nz' in subjacs_info:
deriv['uncovered_nz'] = subjacs_info['uncovered_nz']
deriv['uncovered_threshold'] = subjacs_info['uncovered_threshold']
if _wrt in local_opts and local_opts[_wrt]['directional']:
if self.matrix_free:
# Dot product test for adjoint validity.
m = mfree_directions[_of].flatten()
d = mfree_directions[_wrt].flatten()
mhat = partial.flatten()
dhat = deriv['J_rev'].flatten()
if 'directional_fd_rev' not in deriv:
deriv['directional_fd_rev'] = []
deriv['directional_fd_rev'].append((mhat.dot(m), dhat.dot(d)))
# convert to regular dict from defaultdict
partials_data = {key: dict(d) for key, d in partials_data.items()}
if could_not_cs:
issue_warning(f"Component '{self.pathname}' requested complex step, but "
"force_alloc_complex has not been set to True, so finite difference was "
"used.\nTo enable complex step, specify 'force_alloc_complex=True' when "
"calling setup on the problem, e.g. "
"'problem.setup(force_alloc_complex=True)'.",
category=DerivativesWarning)
incon_keys = self._get_inconsistent_keys()
# if compact_print is True, we'll show all derivatives, even non-dependent ones
nondeps = set() if compact_print else nondep_derivs
# force iterator to run so that error info will be added to partials_data
err_iter = list(_iter_derivs(partials_data, show_only_incorrect, all_fd_options, False,
nondeps, self.matrix_free, abs_err_tol, rel_err_tol,
incon_keys))
worst = None
if out_stream is not None:
if compact_print:
worst = _deriv_display_compact(self, err_iter, partials_data, out_stream,
totals=False,
show_only_incorrect=show_only_incorrect,
show_worst=show_worst)
else:
_deriv_display(self, err_iter, partials_data, rel_err_tol, abs_err_tol, out_stream,
all_fd_options, False, show_only_incorrect)
# check for zero subjacs that are declared as dependent
zero_keys = set()
for key, meta in partials_data.items():
if key in nondep_derivs:
continue
abs_key = rel_key2abs_key(self, key)
if abs_key in self._subjacs_info:
maxmag = max([mag.max() for mag in meta['magnitude']])
if maxmag == 0.0:
zero_keys.add(key)
if zero_keys:
issue_warning(f"\nComponent '{self.pathname}' has zero derivatives for the "
"following variable pairs that were declared as 'dependent': "
f"{sorted(zero_keys)}.\n",
category=DerivativesWarning)
# add pathname to the partials dict to make it compatible with the return value
# from Problem.check_partials and passable to assert_check_partials.
return {self.pathname: partials_data}, worst
def _check_compute_primal_args(self):
"""
Check that the compute_primal method args are in the correct order.
"""
args = list(inspect.signature(self._orig_compute_primal).parameters)
if args and args[0] == 'self':
args = args[1:]
compargs = self._get_compute_primal_argnames()
if args != compargs:
raise RuntimeError(f"{self.msginfo}: compute_primal method args {args} don't match "
f"the args {compargs} mapped from this component's inputs. To "
"map inputs to the compute_primal method, set the name used in "
"compute_primal to the 'primal_name' arg when calling "
"add_input/add_discrete_input. This is only necessary if the "
"declared component input name is not a valid Python name.")
def _check_compute_primal_returns(self):
"""
Check that the compute_primal method returns a tuple.
"""
retnames = get_return_names(self.compute_primal)
if self._valid_name_map:
expected_names = [self._valid_name_map.get(n, n) for n in self._var_rel_names['output']]
expected_names.extend([self._valid_name_map.get(n, n) for n in self._discrete_outputs])
else:
expected_names = list(self._var_rel_names['output'])
expected_names.extend(self._discrete_outputs)
if len(retnames) != len(expected_names):
raise RuntimeError(f"{self.msginfo}: compute_primal method returns {len(retnames)} "
f"values but expected {len(expected_names)}.")
for i, (expname, rname) in enumerate(zip(expected_names, retnames)):
if rname is not None and expname != rname:
raise RuntimeError(f"{self.msginfo}: compute_primal method returns {rname} "
f"for return value {i} but the name of the output that was "
f"mapped for this component is {expname}. To map outputs to "
"the compute_primal method, set the name used in compute_primal "
"to the 'primal_name' arg when calling "
"add_output/add_discrete_output. This is only necessary if "
"the declared component output name is not a valid Python name.")
[docs] def get_declare_partials_calls(self, sparsity=None):
"""
Return a string containing declare_partials() calls based on the subjac sparsity.
Parameters
----------
sparsity : coo_matrix or None
Sparsity matrix to use. If None, compute_sparsity will be called to compute it.
Returns
-------
str
A string containing a declare_partials() call for each nonzero subjac. This
string may be cut and pasted into a component's setup() method.
"""
lines = []
for of, wrt, nzrows, nzcols, _ in self.subjac_sparsity_iter(sparsity=sparsity):
lines.append(f" self.declare_partials(of='{of}', wrt='{wrt}', "
f"rows={list(nzrows)}, cols={list(nzcols)})")
return '\n'.join(lines)
class _DictValues(object):
"""
A dict-like wrapper for a dict of metadata, where getitem returns 'val' from metadata.
"""
def __init__(self, dct):
self._dict = dct
def __getitem__(self, key):
return self._dict[key]['val']
def __setitem__(self, key, value):
self._dict[key]['val'] = value
def __contains__(self, key):
return key in self._dict
def __len__(self):
return len(self._dict)
def __iter__(self):
return iter(self._dict)
def __bool__(self):
return bool(self._dict)
def keys(self):
return self._dict.keys()
def items(self):
return [(key, meta['val']) for key, meta in self._dict.items()]
def values(self):
return [meta['val'] for meta in self._dict.values()]
def set_vals(self, vals):
for key, val in zip(self._dict, vals):
self[key] = val
def _get_fd_options(var, global_method, local_opts, global_step, global_form, global_step_calc,
alloc_complex, global_minimum_step):
local_wrt = var
# Determine if fd or cs.
method = global_method
if local_wrt in local_opts:
local_method = local_opts[local_wrt]['method']
if local_method:
method = local_method
# We can't use CS if we haven't allocated a complex vector, so we fall back on fd.
if method == 'cs' and not alloc_complex:
method = 'fd'
could_not_cs = True
else:
could_not_cs = False
fd_options = {'order': None,
'method': method}
if method == 'cs':
fd_options = ComplexStep.DEFAULT_OPTIONS.copy()
fd_options['method'] = 'cs'
fd_options['form'] = None
fd_options['step_calc'] = None
fd_options['minimum_step'] = None
elif method == 'fd':
fd_options = FiniteDifference.DEFAULT_OPTIONS.copy()
fd_options['method'] = 'fd'
fd_options['form'] = global_form
fd_options['step_calc'] = global_step_calc
fd_options['minimum_step'] = global_minimum_step
if global_step and global_method == method:
fd_options['step'] = global_step
fd_options['directional'] = False
# Precedence: component options > global options > defaults
if local_wrt in local_opts:
for name in ['form', 'step', 'step_calc', 'minimum_step', 'directional']:
value = local_opts[local_wrt][name]
if value is not None:
fd_options[name] = value
return fd_options, could_not_cs
