Source code for ufl.algorithms.apply_derivatives

"""Apply derivatives algorithm which computes the derivatives of a form of expression."""

# Copyright (C) 2008-2016 Martin Sandve Alnæs
#
# This file is part of UFL (https://www.fenicsproject.org)
#
# SPDX-License-Identifier:    LGPL-3.0-or-later

import warnings
from collections import defaultdict
from math import pi

from ufl.action import Action
from ufl.algorithms.analysis import extract_arguments
from ufl.algorithms.map_integrands import map_integrand_dags
from ufl.algorithms.replace_derivative_nodes import replace_derivative_nodes
from ufl.argument import BaseArgument
from ufl.checks import is_cellwise_constant
from ufl.classes import (
    Coefficient,
    ComponentTensor,
    Conj,
    ConstantValue,
    ExprList,
    ExprMapping,
    FloatValue,
    FormArgument,
    Grad,
    Identity,
    Imag,
    Indexed,
    IndexSum,
    JacobianInverse,
    ListTensor,
    Product,
    Real,
    ReferenceGrad,
    ReferenceValue,
    SpatialCoordinate,
    Sum,
    Variable,
    Zero,
)
from ufl.constantvalue import is_true_ufl_scalar, is_ufl_scalar
from ufl.core.base_form_operator import BaseFormOperator
from ufl.core.expr import ufl_err_str
from ufl.core.multiindex import FixedIndex, MultiIndex, indices
from ufl.core.terminal import Terminal
from ufl.corealg.map_dag import map_expr_dag
from ufl.corealg.multifunction import MultiFunction
from ufl.differentiation import (
    BaseFormCoordinateDerivative,
    BaseFormOperatorDerivative,
    CoordinateDerivative,
)
from ufl.domain import extract_unique_domain
from ufl.form import Form, ZeroBaseForm
from ufl.operators import (
    bessel_I,
    bessel_J,
    bessel_K,
    bessel_Y,
    cell_avg,
    conditional,
    cos,
    cosh,
    exp,
    facet_avg,
    ln,
    sign,
    sin,
    sinh,
    sqrt,
)
from ufl.pullback import CustomPullback, PhysicalPullback
from ufl.tensors import as_scalar, as_scalars, as_tensor, unit_indexed_tensor, unwrap_list_tensor

# TODO: Add more rulesets?
# - DivRuleset
# - CurlRuleset
# - ReferenceGradRuleset
# - ReferenceDivRuleset


[docs]class GenericDerivativeRuleset(MultiFunction): """A generic derivative.""" def __init__(self, var_shape): """Initialise.""" MultiFunction.__init__(self) self._var_shape = var_shape # --- Error checking for missing handlers and unexpected types
[docs] def expr(self, o): """Raise error.""" raise ValueError( f"Missing differentiation handler for type {o._ufl_class_.__name__}. " "Have you added a new type?" )
[docs] def unexpected(self, o): """Raise error about unexpected type.""" raise ValueError(f"Unexpected type {o._ufl_class_.__name__} in AD rules.")
[docs] def override(self, o): """Raise error about overriding.""" raise ValueError( f"Type {o._ufl_class_.__name__} must be overridden in specialized AD rule set." )
[docs] def derivative(self, o): """Raise error.""" raise ValueError( f"Unhandled derivative type {o._ufl_class_.__name__}, " "nested differentiation has failed." )
# --- Some types just don't have any derivative, this is just to # --- make algorithm structure generic
[docs] def non_differentiable_terminal(self, o): """Return the non-differentiated object. Labels and indices are not differentiable: it's convenient to return the non-differentiated object. """ return o
label = non_differentiable_terminal multi_index = non_differentiable_terminal # --- Helper functions for creating zeros with the right shapes
[docs] def independent_terminal(self, o): """A zero with correct shape for terminals independent of diff. variable.""" return Zero(o.ufl_shape + self._var_shape)
[docs] def independent_operator(self, o): """A zero with correct shape and indices for operators independent of diff. variable.""" return Zero(o.ufl_shape + self._var_shape, o.ufl_free_indices, o.ufl_index_dimensions)
# --- All derivatives need to define grad and averaging grad = override cell_avg = override facet_avg = override # --- Default rules for terminals # Literals are by definition independent of any differentiation variable constant_value = independent_terminal # Constants are independent of any differentiation constant = independent_terminal # Rules for form arguments must be specified in specialized rule set form_argument = override # Rules for geometric quantities must be specified in specialized rule set geometric_quantity = override # These types are currently assumed independent, but for non-affine domains # this no longer holds and we want to implement rules for them. # facet_normal = independent_terminal # spatial_coordinate = independent_terminal # cell_coordinate = independent_terminal # Measures of cell entities, assuming independent although # this will not be true for all of these for non-affine domains # cell_volume = independent_terminal # circumradius = independent_terminal # facet_area = independent_terminal # cell_surface_area = independent_terminal # min_cell_edge_length = independent_terminal # max_cell_edge_length = independent_terminal # min_facet_edge_length = independent_terminal # max_facet_edge_length = independent_terminal # Other stuff # cell_orientation = independent_terminal # quadrature_weigth = independent_terminal # These types are currently not expected to show up in AD pass. # To make some of these available to the end-user, they need to be # implemented here. # facet_coordinate = unexpected # cell_origin = unexpected # facet_origin = unexpected # cell_facet_origin = unexpected # jacobian = unexpected # jacobian_determinant = unexpected # jacobian_inverse = unexpected # facet_jacobian = unexpected # facet_jacobian_determinant = unexpected # facet_jacobian_inverse = unexpected # cell_facet_jacobian = unexpected # cell_facet_jacobian_determinant = unexpected # cell_facet_jacobian_inverse = unexpected # cell_vertices = unexpected # cell_edge_vectors = unexpected # facet_edge_vectors = unexpected # reference_cell_edge_vectors = unexpected # reference_facet_edge_vectors = unexpected # cell_normal = unexpected # TODO: Expecting rename # cell_normals = unexpected # facet_tangents = unexpected # cell_tangents = unexpected # cell_midpoint = unexpected # facet_midpoint = unexpected # --- Default rules for operators
[docs] def variable(self, o, df, unused_l): """Differentiate a variable.""" return df
# --- Indexing and component handling
[docs] def indexed(self, o, Ap, ii): # TODO: (Partially) duplicated in nesting rules """Differentiate an indexed.""" # Propagate zeros if isinstance(Ap, Zero): return self.independent_operator(o) # Untangle as_tensor(C[kk], jj)[ii] -> C[ll] to simplify # resulting expression if isinstance(Ap, ComponentTensor): B, jj = Ap.ufl_operands if isinstance(B, Indexed): C, kk = B.ufl_operands kk = list(kk) if all(j in kk for j in jj): rep = dict(zip(jj, ii)) Cind = [rep.get(k, k) for k in kk] expr = Indexed(C, MultiIndex(tuple(Cind))) assert expr.ufl_free_indices == o.ufl_free_indices assert expr.ufl_shape == o.ufl_shape return expr # Otherwise a more generic approach r = len(Ap.ufl_shape) - len(ii) if r: kk = indices(r) op = Indexed(Ap, MultiIndex(ii.indices() + kk)) op = as_tensor(op, kk) else: op = Indexed(Ap, ii) return op
[docs] def list_tensor(self, o, *dops): """Differentiate a list_tensor.""" return ListTensor(*dops)
[docs] def component_tensor(self, o, Ap, ii): """Differentiate a component_tensor.""" if isinstance(Ap, Zero): op = self.independent_operator(o) else: Ap, jj = as_scalar(Ap) op = as_tensor(Ap, ii.indices() + jj) return op
# --- Algebra operators
[docs] def index_sum(self, o, Ap, i): """Differentiate an index_sum.""" return IndexSum(Ap, i)
[docs] def sum(self, o, da, db): """Differentiate a sum.""" return da + db
[docs] def product(self, o, da, db): """Differentiate a product.""" # Even though arguments to o are scalar, da and db may be # tensor valued a, b = o.ufl_operands (da, db), ii = as_scalars(da, db) pa = Product(da, b) pb = Product(a, db) s = Sum(pa, pb) if ii: s = as_tensor(s, ii) return s
[docs] def division(self, o, fp, gp): """Differentiate a division.""" f, g = o.ufl_operands if not is_ufl_scalar(f): raise ValueError("Not expecting nonscalar nominator") if not is_true_ufl_scalar(g): raise ValueError("Not expecting nonscalar denominator") # do_df = 1/g # do_dg = -h/g # op = do_df*fp + do_df*gp # op = (fp - o*gp) / g # Get o and gp as scalars, multiply, then wrap as a tensor # again so, oi = as_scalar(o) sgp, gi = as_scalar(gp) o_gp = so * sgp if oi or gi: o_gp = as_tensor(o_gp, oi + gi) op = (fp - o_gp) / g return op
[docs] def power(self, o, fp, gp): """Differentiate a power.""" f, g = o.ufl_operands if not is_true_ufl_scalar(f): raise ValueError("Expecting scalar expression f in f**g.") if not is_true_ufl_scalar(g): raise ValueError("Expecting scalar expression g in f**g.") # Derivation of the general case: o = f(x)**g(x) # do/df = g * f**(g-1) = g / f * o # do/dg = ln(f) * f**g = ln(f) * o # do/df * df + do/dg * dg = o * (g / f * df + ln(f) * dg) if isinstance(gp, Zero): # This probably produces better results for the common # case of f**constant op = fp * g * f ** (g - 1) else: # Note: This produces expressions like (1/w)*w**5 instead of w**4 # op = o * (fp * g / f + gp * ln(f)) # This reuses o op = f ** (g - 1) * ( g * fp + f * ln(f) * gp ) # This gives better accuracy in dolfin integration test # Example: d/dx[x**(x**3)]: # f = x # g = x**3 # df = 1 # dg = 3*x**2 # op1 = o * (fp * g / f + gp * ln(f)) # = x**(x**3) * (x**3/x + 3*x**2*ln(x)) # op2 = f**(g-1) * (g*fp + f*ln(f)*gp) # = x**(x**3-1) * (x**3 + x*3*x**2*ln(x)) return op
[docs] def abs(self, o, df): """Differentiate an abs.""" (f,) = o.ufl_operands # return conditional(eq(f, 0), 0, Product(sign(f), df)) abs is # not complex differentiable, so we workaround the case of a # real F in complex mode by defensively casting to real inside # the sign. return sign(Real(f)) * df
# --- Complex algebra
[docs] def conj(self, o, df): """Differentiate a conj.""" return Conj(df)
[docs] def real(self, o, df): """Differentiate a real.""" return Real(df)
[docs] def imag(self, o, df): """Differentiate a imag.""" return Imag(df)
# --- Mathfunctions
[docs] def math_function(self, o, df): """Differentiate a math_function.""" # FIXME: Introduce a UserOperator type instead of this hack # and define user derivative() function properly if hasattr(o, "derivative"): (f,) = o.ufl_operands return df * o.derivative() raise ValueError("Unknown math function.")
[docs] def sqrt(self, o, fp): """Differentiate a sqrt.""" return fp / (2 * o)
[docs] def exp(self, o, fp): """Differentiate an exp.""" return fp * o
[docs] def ln(self, o, fp): """Differentiate a ln.""" (f,) = o.ufl_operands if isinstance(f, Zero): raise ZeroDivisionError() return fp / f
[docs] def cos(self, o, fp): """Differentiate a cos.""" (f,) = o.ufl_operands return fp * -sin(f)
[docs] def sin(self, o, fp): """Differentiate a sin.""" (f,) = o.ufl_operands return fp * cos(f)
[docs] def tan(self, o, fp): """Differentiate a tan.""" (f,) = o.ufl_operands return 2.0 * fp / (cos(2.0 * f) + 1.0)
[docs] def cosh(self, o, fp): """Differentiate a cosh.""" (f,) = o.ufl_operands return fp * sinh(f)
[docs] def sinh(self, o, fp): """Differentiate a sinh.""" (f,) = o.ufl_operands return fp * cosh(f)
[docs] def tanh(self, o, fp): """Differentiate a tanh.""" (f,) = o.ufl_operands def sech(y): return (2.0 * cosh(y)) / (cosh(2.0 * y) + 1.0) return fp * sech(f) ** 2
[docs] def acos(self, o, fp): """Differentiate an acos.""" (f,) = o.ufl_operands return -fp / sqrt(1.0 - f**2)
[docs] def asin(self, o, fp): """Differentiate an asin.""" (f,) = o.ufl_operands return fp / sqrt(1.0 - f**2)
[docs] def atan(self, o, fp): """Differentiate an atan.""" (f,) = o.ufl_operands return fp / (1.0 + f**2)
[docs] def atan2(self, o, fp, gp): """Differentiate an atan2.""" f, g = o.ufl_operands return (g * fp - f * gp) / (f**2 + g**2)
[docs] def erf(self, o, fp): """Differentiate an erf.""" (f,) = o.ufl_operands return fp * (2.0 / sqrt(pi) * exp(-(f**2)))
# --- Bessel functions
[docs] def bessel_j(self, o, nup, fp): """Differentiate a bessel_j.""" nu, f = o.ufl_operands if not (nup is None or isinstance(nup, Zero)): raise NotImplementedError( "Differentiation of bessel function w.r.t. nu is not supported." ) if isinstance(nu, Zero): op = -bessel_J(1, f) else: op = 0.5 * (bessel_J(nu - 1, f) - bessel_J(nu + 1, f)) return op * fp
[docs] def bessel_y(self, o, nup, fp): """Differentiate a bessel_y.""" nu, f = o.ufl_operands if not (nup is None or isinstance(nup, Zero)): raise NotImplementedError( "Differentiation of bessel function w.r.t. nu is not supported." ) if isinstance(nu, Zero): op = -bessel_Y(1, f) else: op = 0.5 * (bessel_Y(nu - 1, f) - bessel_Y(nu + 1, f)) return op * fp
[docs] def bessel_i(self, o, nup, fp): """Differentiate a bessel_i.""" nu, f = o.ufl_operands if not (nup is None or isinstance(nup, Zero)): raise NotImplementedError( "Differentiation of bessel function w.r.t. nu is not supported." ) if isinstance(nu, Zero): op = bessel_I(1, f) else: op = 0.5 * (bessel_I(nu - 1, f) + bessel_I(nu + 1, f)) return op * fp
[docs] def bessel_k(self, o, nup, fp): """Differentiate a bessel_k.""" nu, f = o.ufl_operands if not (nup is None or isinstance(nup, Zero)): raise NotImplementedError( "Differentiation of bessel function w.r.t. nu is not supported." ) if isinstance(nu, Zero): op = -bessel_K(1, f) else: op = -0.5 * (bessel_K(nu - 1, f) + bessel_K(nu + 1, f)) return op * fp
# --- Restrictions
[docs] def restricted(self, o, fp): """Differentiate a restricted.""" # Restriction and differentiation commutes if isinstance(fp, ConstantValue): return fp # TODO: Add simplification to Restricted instead? else: return fp(o._side) # (f+-)' == (f')+-
# --- Conditionals
[docs] def binary_condition(self, o, dl, dr): """Differentiate a binary_condition.""" # Should not be used anywhere... return None
[docs] def not_condition(self, o, c): """Differentiate a not_condition.""" # Should not be used anywhere... return None
[docs] def conditional(self, o, unused_dc, dt, df): """Differentiate a conditional.""" if isinstance(dt, Zero) and isinstance(df, Zero): # Assuming dt and df have the same indices here, which # should be the case return dt else: # Not placing t[1],f[1] outside, allowing arguments inside # conditionals. This will make legacy ffc fail, but # should work with uflacs. c = o.ufl_operands[0] return conditional(c, dt, df)
[docs] def max_value(self, o, df, dg): """Differentiate a max_value.""" # d/dx max(f, g) = # f > g: df/dx # f < g: dg/dx # Placing df,dg outside here to avoid getting arguments inside # conditionals f, g = o.ufl_operands dc = conditional(f > g, 1, 0) return dc * df + (1.0 - dc) * dg
[docs] def min_value(self, o, df, dg): """Differentiate a min_value.""" # d/dx min(f, g) = # f < g: df/dx # else: dg/dx # Placing df,dg outside here to avoid getting arguments # inside conditionals f, g = o.ufl_operands dc = conditional(f < g, 1, 0) return dc * df + (1.0 - dc) * dg
[docs]class GradRuleset(GenericDerivativeRuleset): """Take the grad derivative.""" def __init__(self, geometric_dimension): """Initialise.""" GenericDerivativeRuleset.__init__(self, var_shape=(geometric_dimension,)) self._Id = Identity(geometric_dimension) # --- Specialized rules for geometric quantities
[docs] def geometric_quantity(self, o): """Differentiate a geometric_quantity. Default for geometric quantities is do/dx = 0 if piecewise constant, otherwise transform derivatives to reference derivatives. Override for specific types if other behaviour is needed. """ if is_cellwise_constant(o): return self.independent_terminal(o) else: domain = extract_unique_domain(o) K = JacobianInverse(domain) Do = grad_to_reference_grad(o, K) return Do
[docs] def jacobian_inverse(self, o): """Differentiate a jacobian_inverse.""" # grad(K) == K_ji rgrad(K)_rj if is_cellwise_constant(o): return self.independent_terminal(o) if not o._ufl_is_terminal_: raise ValueError("ReferenceValue can only wrap a terminal") Do = grad_to_reference_grad(o, o) return Do
# TODO: Add more explicit geometry type handlers here, with # non-affine domains several should be non-zero.
[docs] def spatial_coordinate(self, o): """Differentiate a spatial_coordinate. dx/dx = I. """ return self._Id
[docs] def cell_coordinate(self, o): """Differentiate a cell_coordinate. dX/dx = inv(dx/dX) = inv(J) = K. """ # FIXME: Is this true for manifolds? What about orientation? return JacobianInverse(extract_unique_domain(o))
# --- Specialized rules for form arguments
[docs] def base_form_operator(self, o): """Differentiate a base_form_operator.""" # Push the grad through the operator is not legal in most cases: # -> Not enouth regularity for chain rule to hold! # By the time we evaluate `grad(o)`, the operator `o` will have # been assembled and substituted by its output. return Grad(o)
[docs] def coefficient(self, o): """Differentiate a coefficient.""" if is_cellwise_constant(o): return self.independent_terminal(o) return Grad(o)
[docs] def argument(self, o): """Differentiate an argument.""" # TODO: Enable this after fixing issue#13, unless we move # simplificat ion to a separate stage? # if is_cellwise_constant(o): # # Collapse gradient of cellwise constant function to zero # # TODO: Missing this type # return AnnotatedZero(o.ufl_shape + self._var_shape, arguments=(o,)) return Grad(o)
# --- Rules for values or derivatives in reference frame
[docs] def reference_value(self, o): """Differentiate a reference_value.""" # grad(o) == grad(rv(f)) -> K_ji*rgrad(rv(f))_rj f = o.ufl_operands[0] if isinstance(f.ufl_element().pullback, PhysicalPullback): # TODO: Do we need to be more careful for immersed things? return ReferenceGrad(o) if not f._ufl_is_terminal_: raise ValueError("ReferenceValue can only wrap a terminal") domain = extract_unique_domain(f) K = JacobianInverse(domain) Do = grad_to_reference_grad(o, K) return Do
[docs] def reference_grad(self, o): """Differentiate a reference_grad.""" # grad(o) == grad(rgrad(rv(f))) -> K_ji*rgrad(rgrad(rv(f)))_rj f = o.ufl_operands[0] valid_operand = f._ufl_is_in_reference_frame_ or isinstance( f, (JacobianInverse, SpatialCoordinate) ) if not valid_operand: raise ValueError("ReferenceGrad can only wrap a reference frame type!") domain = extract_unique_domain(f) K = JacobianInverse(domain) Do = grad_to_reference_grad(o, K) return Do
# --- Nesting of gradients
[docs] def grad(self, o): """Differentiate a grad. Represent grad(grad(f)) as Grad(Grad(f)). """ # Check that o is a "differential terminal" if not isinstance(o.ufl_operands[0], (Grad, Terminal)): raise ValueError("Expecting only grads applied to a terminal.") return Grad(o)
def _grad(self, o): """Differentiate a _grad.""" pass # TODO: Not sure how to detect that gradient of f is cellwise constant. # Can we trust element degrees? # if is_cellwise_constant(o): # return self.terminal(o) # TODO: Maybe we can ask "f.has_derivatives_of_order(n)" to check # if we should make a zero here? # 1) n = count number of Grads, get f # 2) if not f.has_derivatives(n): return zero(...) cell_avg = GenericDerivativeRuleset.independent_operator facet_avg = GenericDerivativeRuleset.independent_operator
[docs]def grad_to_reference_grad(o, K): """Relates grad(o) to reference_grad(o) using the Jacobian inverse. Args: o: Operand K: Jacobian inverse Returns: grad(o) written in terms of reference_grad(o) and K """ r = indices(len(o.ufl_shape)) i, j = indices(2) # grad(o) == K_ji rgrad(o)_rj Do = as_tensor(K[j, i] * ReferenceGrad(o)[r + (j,)], r + (i,)) return Do
[docs]class ReferenceGradRuleset(GenericDerivativeRuleset): """Apply the reference grad derivative.""" def __init__(self, topological_dimension): """Initialise.""" GenericDerivativeRuleset.__init__(self, var_shape=(topological_dimension,)) self._Id = Identity(topological_dimension) # --- Specialized rules for geometric quantities
[docs] def geometric_quantity(self, o): """Differentiate a geometric_quantity. dg/dX = 0 if piecewise constant, otherwise ReferenceGrad(g). """ if is_cellwise_constant(o): return self.independent_terminal(o) else: # TODO: Which types does this involve? I don't think the # form compilers will handle this. return ReferenceGrad(o)
[docs] def spatial_coordinate(self, o): """Differentiate a spatial_coordinate. dx/dX = J. """ # Don't convert back to J, otherwise we get in a loop return ReferenceGrad(o)
[docs] def cell_coordinate(self, o): """Differentiate a cell_coordinate. dX/dX = I. """ return self._Id
# TODO: Add more geometry types here, with non-affine domains # several should be non-zero. # --- Specialized rules for form arguments
[docs] def reference_value(self, o): """Differentiate a reference_value.""" if not o.ufl_operands[0]._ufl_is_terminal_: raise ValueError("ReferenceValue can only wrap a terminal") return ReferenceGrad(o)
[docs] def coefficient(self, o): """Differentiate a coefficient.""" raise ValueError("Coefficient should be wrapped in ReferenceValue by now")
[docs] def argument(self, o): """Differentiate an argument.""" raise ValueError("Argument should be wrapped in ReferenceValue by now")
# --- Nesting of gradients
[docs] def grad(self, o): """Differentiate a grad.""" raise ValueError( f"Grad should have been transformed by this point, but got {type(o).__name__}." )
[docs] def reference_grad(self, o): """Differentiate a reference_grad. Represent ref_grad(ref_grad(f)) as RefGrad(RefGrad(f)). """ # Check that o is a "differential terminal" if not isinstance(o.ufl_operands[0], (ReferenceGrad, ReferenceValue, Terminal)): raise ValueError("Expecting only grads applied to a terminal.") return ReferenceGrad(o)
cell_avg = GenericDerivativeRuleset.independent_operator facet_avg = GenericDerivativeRuleset.independent_operator
[docs]class VariableRuleset(GenericDerivativeRuleset): """Differentiate with respect to a variable.""" def __init__(self, var): """Initialise.""" GenericDerivativeRuleset.__init__(self, var_shape=var.ufl_shape) if var.ufl_free_indices: raise ValueError("Differentiation variable cannot have free indices.") self._variable = var self._Id = self._make_identity(self._var_shape) def _make_identity(self, sh): """Differentiate a _make_identity. Creates a higher order identity tensor to represent dv/dv. """ res = None if sh == (): # Scalar dv/dv is scalar return FloatValue(1.0) elif len(sh) == 1: # Vector v makes dv/dv the identity matrix return Identity(sh[0]) else: # TODO: Add a type for this higher order identity? # II[i0,i1,i2,j0,j1,j2] = 1 if all((i0==j0, i1==j1, i2==j2)) else 0 # Tensor v makes dv/dv some kind of higher rank identity tensor ind1 = () ind2 = () for d in sh: i, j = indices(2) dij = Identity(d)[i, j] if res is None: res = dij else: res *= dij ind1 += (i,) ind2 += (j,) fp = as_tensor(res, ind1 + ind2) return fp # Explicitly defining dg/dw == 0 geometric_quantity = GenericDerivativeRuleset.independent_terminal # Explicitly defining da/dw == 0 argument = GenericDerivativeRuleset.independent_terminal
[docs] def coefficient(self, o): """Differentiate a coefficient. df/dv = Id if v is f else 0. Note that if v = variable(f), df/dv is still 0, but if v == f, i.e. isinstance(v, Coefficient) == True, then df/dv == df/df = Id. """ v = self._variable if isinstance(v, Coefficient) and o == v: # dv/dv = identity of rank 2*rank(v) return self._Id else: # df/v = 0 return self.independent_terminal(o)
[docs] def variable(self, o, df, a): """Differentiate a variable.""" v = self._variable if isinstance(v, Variable) and v.label() == a: # dv/dv = identity of rank 2*rank(v) return self._Id else: # df/v = df return df
[docs] def grad(self, o): """Differentiate a grad. Variable derivative of a gradient of a terminal must be 0. """ # Check that o is a "differential terminal" if not isinstance(o.ufl_operands[0], (Grad, Terminal)): raise ValueError("Expecting only grads applied to a terminal.") return self.independent_terminal(o)
# --- Rules for values or derivatives in reference frame
[docs] def reference_value(self, o): """Differentiate a reference_value.""" # d/dv(o) == d/dv(rv(f)) = 0 if v is not f, or rv(dv/df) v = self._variable if isinstance(v, Coefficient) and o.ufl_operands[0] == v: if not v.ufl_element().pullback.is_identity: # FIXME: This is a bit tricky, instead of Identity it is # actually inverse(transform), or we should rather not # convert to reference frame in the first place raise ValueError( "Missing implementation: To handle derivatives of rv(f) w.r.t. f for " "mapped elements, rewriting to reference frame should not happen first..." ) # dv/dv = identity of rank 2*rank(v) return self._Id else: # df/v = 0 return self.independent_terminal(o)
[docs] def reference_grad(self, o): """Differentiate a reference_grad. Variable derivative of a gradient of a terminal must be 0. """ if not isinstance(o.ufl_operands[0], (ReferenceGrad, ReferenceValue)): raise ValueError("Unexpected argument to reference_grad.") return self.independent_terminal(o)
cell_avg = GenericDerivativeRuleset.independent_operator facet_avg = GenericDerivativeRuleset.independent_operator
[docs]class GateauxDerivativeRuleset(GenericDerivativeRuleset): """Apply AFD (Automatic Functional Differentiation) to expression. Implements rules for the Gateaux derivative D_w[v](...) defined as D_w[v](e) = d/dtau e(w+tau v)|tau=0. """ def __init__(self, coefficients, arguments, coefficient_derivatives, pending_operations): """Initialise.""" GenericDerivativeRuleset.__init__(self, var_shape=()) # Type checking if not isinstance(coefficients, ExprList): raise ValueError("Expecting a ExprList of coefficients.") if not isinstance(arguments, ExprList): raise ValueError("Expecting a ExprList of arguments.") if not isinstance(coefficient_derivatives, ExprMapping): raise ValueError("Expecting a coefficient-coefficient ExprMapping.") # The coefficient(s) to differentiate w.r.t. and the # argument(s) s.t. D_w[v](e) = d/dtau e(w+tau v)|tau=0 self._w = coefficients.ufl_operands self._v = arguments.ufl_operands self._w2v = {w: v for w, v in zip(self._w, self._v)} # Build more convenient dict {f: df/dw} for each coefficient f # where df/dw is nonzero cd = coefficient_derivatives.ufl_operands self._cd = {cd[2 * i]: cd[2 * i + 1] for i in range(len(cd) // 2)} # Record the operations delayed to the derivative expansion phase: # Example: dN(u)/du where `N` is an ExternalOperator and `u` a Coefficient self.pending_operations = pending_operations # Explicitly defining dg/dw == 0 geometric_quantity = GenericDerivativeRuleset.independent_terminal
[docs] def cell_avg(self, o, fp): """Differentiate a cell_avg.""" # Cell average of a single function and differentiation # commutes, D_f[v](cell_avg(f)) = cell_avg(v) return cell_avg(fp)
[docs] def facet_avg(self, o, fp): """Differentiate a facet_avg.""" # Facet average of a single function and differentiation # commutes, D_f[v](facet_avg(f)) = facet_avg(v) return facet_avg(fp)
# Explicitly defining da/dw == 0 argument = GenericDerivativeRuleset.independent_terminal
[docs] def coefficient(self, o): """Differentiate a coefficient.""" # Define dw/dw := d/ds [w + s v] = v # Return corresponding argument if we can find o among w do = self._w2v.get(o) if do is not None: return do # Look for o among coefficient derivatives dos = self._cd.get(o) if dos is None: # If o is not among coefficient derivatives, return # do/dw=0 do = Zero(o.ufl_shape) return do else: # Compute do/dw_j = do/dw_h : v. # Since we may actually have a tuple of oprimes and vs in a # 'mixed' space, sum over them all to get the complete inner # product. Using indices to define a non-compound inner product. # Example: # (f:g) -> (dfdu:v):g + f:(dgdu:v) # shape(dfdu) == shape(f) + shape(v) # shape(f) == shape(g) == shape(dfdu : v) # Make sure we have a tuple to match the self._v tuple if not isinstance(dos, tuple): dos = (dos,) if len(dos) != len(self._v): raise ValueError( "Got a tuple of arguments, expecting a " "matching tuple of coefficient derivatives." ) dosum = Zero(o.ufl_shape) for do, v in zip(dos, self._v): so, oi = as_scalar(do) rv = len(oi) - len(v.ufl_shape) oi1 = oi[:rv] oi2 = oi[rv:] prod = so * v[oi2] if oi1: dosum += as_tensor(prod, oi1) else: dosum += prod return dosum
[docs] def reference_value(self, o): """Differentiate a reference_value.""" raise NotImplementedError( "Currently no support for ReferenceValue in CoefficientDerivative." )
# TODO: This is implementable for regular derivative(M(f),f,v) # but too messy if customized coefficient derivative # relations are given by the user. We would only need # this to allow the user to write # derivative(...ReferenceValue...,...). # f, = o.ufl_operands # if not f._ufl_is_terminal_: # raise ValueError("ReferenceValue can only wrap terminals directly.") # FIXME: check all cases like in coefficient # if f is w: # # FIXME: requires that v is an Argument with the same element mapping! # return ReferenceValue(v) # else: # return self.independent_terminal(o)
[docs] def reference_grad(self, o): """Differentiate a reference_grad.""" raise NotImplementedError( "Currently no support for ReferenceGrad in CoefficientDerivative." )
# TODO: This is implementable for regular derivative(M(f),f,v) # but too messy if customized coefficient derivative # relations are given by the user. We would only need # this to allow the user to write # derivative(...ReferenceValue...,...).
[docs] def grad(self, g): """Differentiate a grad.""" # If we hit this type, it has already been propagated to a # coefficient (or grad of a coefficient) or a base form operator, # FIXME: Assert # this! so we need to take the gradient of the variation or # return zero. Complications occur when dealing with # derivatives w.r.t. single components... # Figure out how many gradients are around the inner terminal ngrads = 0 o = g while isinstance(o, Grad): (o,) = o.ufl_operands ngrads += 1 # `grad(N)` where N is a BaseFormOperator is treated as if `N` was a Coefficient. if not isinstance(o, (FormArgument, BaseFormOperator)): raise ValueError(f"Expecting gradient of a FormArgument, not {ufl_err_str(o)}.") def apply_grads(f): for i in range(ngrads): f = Grad(f) return f # Find o among all w without any indexing, which makes this # easy for w, v in zip(self._w, self._v): if o == w and isinstance(v, FormArgument): # Case: d/dt [w + t v] return apply_grads(v) # If o is not among coefficient derivatives, return do/dw=0 gprimesum = Zero(g.ufl_shape) def analyse_variation_argument(v): # Analyse variation argument if isinstance(v, FormArgument): # Case: d/dt [w[...] + t v] vval, vcomp = v, () elif isinstance(v, Indexed): # Case: d/dt [w + t v[...]] # Case: d/dt [w[...] + t v[...]] vval, vcomp = v.ufl_operands vcomp = tuple(vcomp) else: raise ValueError("Expecting argument or component of argument.") if not all(isinstance(k, FixedIndex) for k in vcomp): raise ValueError("Expecting only fixed indices in variation.") return vval, vcomp def compute_gprimeterm(ngrads, vval, vcomp, wshape, wcomp): # Apply gradients directly to argument vval, and get the # right indexed scalar component(s) kk = indices(ngrads) Dvkk = apply_grads(vval)[vcomp + kk] # Place scalar component(s) Dvkk into the right tensor # positions if wshape: Ejj, jj = unit_indexed_tensor(wshape, wcomp) else: Ejj, jj = 1, () gprimeterm = as_tensor(Ejj * Dvkk, jj + kk) return gprimeterm # Accumulate contributions from variations in different # components for w, v in zip(self._w, self._v): # -- Analyse differentiation variable coefficient -- # # Can differentiate a Form wrt a BaseFormOperator if isinstance(w, (FormArgument, BaseFormOperator)): if not w == o: continue wshape = w.ufl_shape if isinstance(v, FormArgument): # Case: d/dt [w + t v] return apply_grads(v) elif isinstance(v, ListTensor): # Case: d/dt [w + t <...,v,...>] for wcomp, vsub in unwrap_list_tensor(v): if not isinstance(vsub, Zero): vval, vcomp = analyse_variation_argument(vsub) gprimesum = gprimesum + compute_gprimeterm( ngrads, vval, vcomp, wshape, wcomp ) else: if wshape != (): raise ValueError("Expecting scalar coefficient in this branch.") # Case: d/dt [w + t v[...]] wval, wcomp = w, () vval, vcomp = analyse_variation_argument(v) gprimesum = gprimesum + compute_gprimeterm(ngrads, vval, vcomp, wshape, wcomp) elif isinstance( w, Indexed ): # This path is tested in unit tests, but not actually used? # Case: d/dt [w[...] + t v[...]] # Case: d/dt [w[...] + t v] wval, wcomp = w.ufl_operands if not wval == o: continue assert isinstance(wval, FormArgument) if not all(isinstance(k, FixedIndex) for k in wcomp): raise ValueError("Expecting only fixed indices in differentiation variable.") wshape = wval.ufl_shape vval, vcomp = analyse_variation_argument(v) gprimesum = gprimesum + compute_gprimeterm(ngrads, vval, vcomp, wshape, wcomp) else: raise ValueError("Expecting coefficient or component of coefficient.") # FIXME: Handle other coefficient derivatives: oprimes = # self._cd.get(o) if 0: oprimes = self._cd.get(o) if oprimes is None: if self._cd: # TODO: Make it possible to silence this message # in particular? It may be good to have for # debugging... warnings.warn(f"Assuming d{{{0}}}/d{{{self._w}}} = 0.") else: # Make sure we have a tuple to match the self._v tuple if not isinstance(oprimes, tuple): oprimes = (oprimes,) if len(oprimes) != len(self._v): raise ValueError( "Got a tuple of arguments, expecting a" " matching tuple of coefficient derivatives." ) # Compute dg/dw_j = dg/dw_h : v. # Since we may actually have a tuple of oprimes and vs # in a 'mixed' space, sum over them all to get the # complete inner product. Using indices to define a # non-compound inner product. for oprime, v in zip(oprimes, self._v): raise NotImplementedError("FIXME: Figure out how to do this with ngrads") so, oi = as_scalar(oprime) rv = len(v.ufl_shape) oi1 = oi[:-rv] oi2 = oi[-rv:] prod = so * v[oi2] if oi1: gprimesum += as_tensor(prod, oi1) else: gprimesum += prod return gprimesum
[docs] def coordinate_derivative(self, o): """Differentiate a coordinate_derivative.""" o = o.ufl_operands return CoordinateDerivative(map_expr_dag(self, o[0]), o[1], o[2], o[3])
[docs] def base_form_operator(self, o, *dfs): """Differentiate a base_form_operator. If d_coeff = 0 => BaseFormOperator's derivative is taken wrt a variable => we call the appropriate handler. Otherwise => differentiation done wrt the BaseFormOperator (dF/dN[Nhat]) => we treat o as a Coefficient. """ d_coeff = self.coefficient(o) # It also handles the non-scalar case if d_coeff == 0: self.pending_operations += (o,) return d_coeff
# -- Handlers for BaseForm objects -- #
[docs] def cofunction(self, o): """Differentiate a cofunction.""" # Same rule than for Coefficient except that we use a Coargument. # The coargument is already attached to the class (self._v) # which `self.coefficient` relies on. dc = self.coefficient(o) if dc == 0: # Convert ufl.Zero into ZeroBaseForm return ZeroBaseForm(o.arguments() + self._v) return dc
[docs] def coargument(self, o): """Differentiate a coargument.""" # Same rule than for Argument (da/dw == 0). dc = self.argument(o) if dc == 0: # Convert ufl.Zero into ZeroBaseForm return ZeroBaseForm(o.arguments() + self._v) return dc
[docs] def matrix(self, M): """Differentiate a matrix.""" # Matrix rule: D_w[v](M) = v if M == w else 0 # We can't differentiate wrt a matrix so always return zero in # the appropriate space return ZeroBaseForm(M.arguments() + self._v)
[docs]class BaseFormOperatorDerivativeRuleset(GateauxDerivativeRuleset): """Apply AFD (Automatic Functional Differentiation) to BaseFormOperator. Implements rules for the Gateaux derivative D_w[v](...) defined as D_w[v](B) = d/dtau B(w+tau v)|tau=0 where B is a ufl.BaseFormOperator. """ def __init__(self, coefficients, arguments, coefficient_derivatives, pending_operations): """Initialise.""" GateauxDerivativeRuleset.__init__( self, coefficients, arguments, coefficient_derivatives, pending_operations )
[docs] def pending_operations_recording(base_form_operator_handler): """Decorate a function to record pending operations.""" def wrapper(self, base_form_op, *dfs): """Decorate.""" # Get the outer `BaseFormOperator` expression, i.e. the # operator that is being differentiated. expression = self.pending_operations.expression # If the base form operator we observe is different from the # outer `BaseFormOperator`: # -> Record that `BaseFormOperator` so that # `d(expression)/d(base_form_op)` can then be computed # later. # Else: # -> Compute the Gateaux derivative of `base_form_ops` by # calling the appropriate handler. if expression != base_form_op: self.pending_operations += (base_form_op,) return self.coefficient(base_form_op) return base_form_operator_handler(self, base_form_op, *dfs) return wrapper
[docs] @pending_operations_recording def interpolate(self, i_op, dw): """Differentiate an interpolate.""" # Interpolate rule: D_w[v](i_op(w, v*)) = i_op(v, v*), by linearity of Interpolate! if not dw: # i_op doesn't depend on w: # -> It also covers the Hessian case since Interpolate is linear, # e.g. D_w[v](D_w[v](i_op(w, v*))) = D_w[v](i_op(v, v*)) = 0 (since w not found). return ZeroBaseForm(i_op.arguments() + self._v) return i_op._ufl_expr_reconstruct_(expr=dw)
[docs] @pending_operations_recording def external_operator(self, N, *dfs): """Differentiate an external_operator.""" result = () for i, df in enumerate(dfs): derivatives = tuple(dj + int(i == j) for j, dj in enumerate(N.derivatives)) if len(extract_arguments(df)) != 0: # Handle the symbolic differentiation of external operators. # This bit returns: # # `\sum_{i} dNdOi(..., Oi, ...; DOi(u)[v], ..., v*)` # # where we differentate wrt u, Oi is the i-th operand, # N(..., Oi, ...; ..., v*) an ExternalOperator and v the # direction (Argument). dNdOi(..., Oi, ...; DOi(u)[v]) # is an ExternalOperator representing the # Gateaux-derivative of N. For example: # -> From N(u) = u**2, we get `dNdu(u; uhat, v*) = 2 * u * uhat`. new_args = N.argument_slots() + (df,) extop = N._ufl_expr_reconstruct_( *N.ufl_operands, derivatives=derivatives, argument_slots=new_args ) elif df == 0: extop = Zero(N.ufl_shape) else: raise NotImplementedError( "Frechet derivative of external operators need to be provided!" ) result += (extop,) return sum(result)
[docs]class DerivativeRuleDispatcher(MultiFunction): """Dispatch a derivative rule.""" def __init__(self): """Initialise.""" MultiFunction.__init__(self) # caches for reuse in the dispatched transformers self.vcaches = defaultdict(dict) self.rcaches = defaultdict(dict) # Record the operations delayed to the derivative expansion phase: # Example: dN(u)/du where `N` is a BaseFormOperator and `u` a Coefficient self.pending_operations = ()
[docs] def terminal(self, o): """Apply to a terminal.""" return o
[docs] def derivative(self, o): """Apply to a derivative.""" raise NotImplementedError(f"Missing derivative handler for {type(o).__name__}.")
ufl_type = MultiFunction.reuse_if_untouched
[docs] def grad(self, o, f): """Apply to a grad.""" rules = GradRuleset(o.ufl_shape[-1]) key = (GradRuleset, o.ufl_shape[-1]) return map_expr_dag(rules, f, vcache=self.vcaches[key], rcache=self.rcaches[key])
[docs] def reference_grad(self, o, f): """Apply to a reference_grad.""" rules = ReferenceGradRuleset(o.ufl_shape[-1]) # FIXME: Look over this and test better. key = (ReferenceGradRuleset, o.ufl_shape[-1]) return map_expr_dag(rules, f, vcache=self.vcaches[key], rcache=self.rcaches[key])
[docs] def variable_derivative(self, o, f, dummy_v): """Apply to a variable_derivative.""" op = o.ufl_operands[1] rules = VariableRuleset(op) key = (VariableRuleset, op) return map_expr_dag(rules, f, vcache=self.vcaches[key], rcache=self.rcaches[key])
[docs] def coefficient_derivative(self, o, f, dummy_w, dummy_v, dummy_cd): """Apply to a coefficient_derivative.""" dummy, w, v, cd = o.ufl_operands pending_operations = BaseFormOperatorDerivativeRecorder( f, w, arguments=v, coefficient_derivatives=cd ) rules = GateauxDerivativeRuleset(w, v, cd, pending_operations) key = (GateauxDerivativeRuleset, w, v, cd) # We need to go through the dag first to record the pending # operations mapped_expr = map_expr_dag(rules, f, vcache=self.vcaches[key], rcache=self.rcaches[key]) # Need to account for pending operations that have been stored # in other integrands self.pending_operations += pending_operations return mapped_expr
[docs] def base_form_operator_derivative(self, o, f, dummy_w, dummy_v, dummy_cd): """Apply to a base_form_operator_derivative.""" dummy, w, v, cd = o.ufl_operands pending_operations = BaseFormOperatorDerivativeRecorder( f, w, arguments=v, coefficient_derivatives=cd ) rules = BaseFormOperatorDerivativeRuleset(w, v, cd, pending_operations=pending_operations) key = (BaseFormOperatorDerivativeRuleset, w, v, cd) if isinstance(f, ZeroBaseForm): (arg,) = v.ufl_operands arguments = f.arguments() # derivative(F, u, du) with `du` a Coefficient # is equivalent to taking the action of the derivative. # In that case, we don't add arguments to `ZeroBaseForm`. if isinstance(arg, BaseArgument): arguments += (arg,) return ZeroBaseForm(arguments) # We need to go through the dag first to record the pending operations mapped_expr = map_expr_dag(rules, f, vcache=self.vcaches[key], rcache=self.rcaches[key]) mapped_f = rules.coefficient(f) if mapped_f != 0: # If dN/dN needs to return an Argument in N space # with N a BaseFormOperator. return mapped_f # Need to account for pending operations that have been stored in other integrands self.pending_operations += pending_operations return mapped_expr
[docs] def coordinate_derivative(self, o, f, dummy_w, dummy_v, dummy_cd): """Apply to a coordinate_derivative.""" o_ = o.ufl_operands key = (CoordinateDerivative, o_[0]) return CoordinateDerivative( map_expr_dag(self, o_[0], vcache=self.vcaches[key], rcache=self.rcaches[key]), o_[1], o_[2], o_[3], )
[docs] def base_form_coordinate_derivative(self, o, f, dummy_w, dummy_v, dummy_cd): """Apply to a base_form_coordinate_derivative.""" o_ = o.ufl_operands key = (BaseFormCoordinateDerivative, o_[0]) return BaseFormCoordinateDerivative( map_expr_dag(self, o_[0], vcache=self.vcaches[key], rcache=self.rcaches[key]), o_[1], o_[2], o_[3], )
[docs] def indexed(self, o, Ap, ii): # TODO: (Partially) duplicated in generic rules """Apply to an indexed.""" # Reuse if untouched if Ap is o.ufl_operands[0]: return o # Untangle as_tensor(C[kk], jj)[ii] -> C[ll] to simplify # resulting expression if isinstance(Ap, ComponentTensor): B, jj = Ap.ufl_operands if isinstance(B, Indexed): C, kk = B.ufl_operands kk = list(kk) if all(j in kk for j in jj): rep = dict(zip(jj, ii)) Cind = [rep.get(k, k) for k in kk] expr = Indexed(C, MultiIndex(tuple(Cind))) assert expr.ufl_free_indices == o.ufl_free_indices assert expr.ufl_shape == o.ufl_shape return expr # Otherwise a more generic approach r = len(Ap.ufl_shape) - len(ii) if r: kk = indices(r) op = Indexed(Ap, MultiIndex(ii.indices() + kk)) op = as_tensor(op, kk) else: op = Indexed(Ap, ii) return op
[docs]class BaseFormOperatorDerivativeRecorder: """A derivative recorded for a base form operator.""" def __init__(self, expression, var, **kwargs): """Initialise.""" base_form_ops = kwargs.pop("base_form_ops", ()) if kwargs.keys() != {"arguments", "coefficient_derivatives"}: raise ValueError( "Only `arguments` and `coefficient_derivatives` are " "allowed as derivative arguments." ) self.expression = expression self.var = var self.der_kwargs = kwargs self.base_form_ops = base_form_ops def __len__(self): """Get the length.""" return len(self.base_form_ops) def __bool__(self): """Convert to a bool.""" return bool(self.base_form_ops) def __add__(self, other): """Add.""" if isinstance(other, (list, tuple)): base_form_ops = self.base_form_ops + other elif isinstance(other, BaseFormOperatorDerivativeRecorder): if self.der_kwargs != other.der_kwargs: raise ValueError( f"Derivative arguments must match when summing {type(self).__name__} objects." ) base_form_ops = self.base_form_ops + other.base_form_ops else: raise NotImplementedError( f"Sum of {type(self)} and {type(other)} objects is not supported." ) return BaseFormOperatorDerivativeRecorder( self.expression, self.var, base_form_ops=base_form_ops, **self.der_kwargs ) def __radd__(self, other): """Add.""" # Recording order doesn't matter as collected # `BaseFormOperator`s are sorted later on. return self.__add__(other) def __iadd__(self, other): """Add.""" if isinstance(other, (list, tuple)): self.base_form_ops += other elif isinstance(other, BaseFormOperatorDerivativeRecorder): self.base_form_ops += other.base_form_ops else: raise NotImplementedError return self
[docs]def apply_derivatives(expression): """Apply derivatives to an expression. Args: expression: A Form, an Expr or a BaseFormOperator to be differentiated Returns: A differentiated expression """ # Notation: Let `var` be the thing we are differentating with respect to. rules = DerivativeRuleDispatcher() # If we hit a base form operator (bfo), then if `var` is: # - a BaseFormOperator → Return `d(expression)/dw` where `w` is # the coefficient produced by the bfo `var`. # - else → Record the bfo on the MultiFunction object and returns # - 0. # Example: # → If derivative(F(u, N(u); v), u) was taken the following line would compute `∂F/∂u`. dexpression_dvar = map_integrand_dags(rules, expression) # Get the recorded delayed operations pending_operations = rules.pending_operations if not pending_operations: return dexpression_dvar # Don't take into account empty Forms if not (isinstance(dexpression_dvar, Form) and len(dexpression_dvar.integrals()) == 0): dexpression_dvar = (dexpression_dvar,) else: dexpression_dvar = () # Retrieve the base form operators, var, and the argument and # coefficient_derivatives for `derivative` var = pending_operations.var base_form_ops = pending_operations.base_form_ops der_kwargs = pending_operations.der_kwargs for N in sorted(set(base_form_ops), key=lambda x: x.count()): # -- Replace dexpr/dvar by dexpr/dN -- # # We don't use `apply_derivatives` since the differentiation is # done via `\partial` and not `d`. dexpr_dN = map_integrand_dags( rules, replace_derivative_nodes(expression, {var.ufl_operands[0]: N}) ) # -- Add the BaseFormOperatorDerivative node -- # (var_arg,) = der_kwargs["arguments"].ufl_operands cd = der_kwargs["coefficient_derivatives"] # Not always the case since `derivative`'s syntax enables one to # use a Coefficient as the Gateaux direction if isinstance(var_arg, BaseArgument): # Construct the argument number based on the # BaseFormOperator arguments instead of naively using # `var_arg`. This is critical when BaseFormOperators are # used inside 0-forms. # # Example: F = 0.5 * u** 2 * dx + 0.5 * N(u; v*)** 2 * dx # -> dFdu[vhat] = <u, vhat> + Action(<N(u; v*), v0>, dNdu(u; v1, v*)) # with `vhat` a 0-numbered argument, and where `v1` and # `vhat` have the same function space but a different # number. Here, applying `vhat` (`var_arg`) naively would # result in `dNdu(u; vhat, v*)`, i.e. the 2-forms `dNdu` # would have two 0-numbered arguments. Instead we increment # the argument number of `vhat` to form `v1`. var_arg = type(var_arg)( var_arg.ufl_function_space(), number=len(N.arguments()), part=var_arg.part() ) dN_dvar = apply_derivatives(BaseFormOperatorDerivative(N, var, ExprList(var_arg), cd)) # -- Sum the Action: dF/du = ∂F/∂u + \sum_{i=1,...} Action(∂F/∂Ni, dNi/du) -- # if not (isinstance(dexpr_dN, Form) and len(dexpr_dN.integrals()) == 0): # In this case: Action <=> ufl.action since `dN_var` has 2 arguments. # We use Action to handle the trivial case `dN_dvar` = 0. dexpression_dvar += (Action(dexpr_dN, dN_dvar),) return sum(dexpression_dvar)
[docs]class CoordinateDerivativeRuleset(GenericDerivativeRuleset): """Apply AFD (Automatic Functional Differentiation) to expression. Implements rules for the Gateaux derivative D_w[v](...) defined as D_w[v](e) = d/dtau e(w+tau v)|tau=0 where 'e' is a ufl form after pullback and w is a SpatialCoordinate. """ def __init__(self, coefficients, arguments, coefficient_derivatives): """Initialise.""" GenericDerivativeRuleset.__init__(self, var_shape=()) # Type checking if not isinstance(coefficients, ExprList): raise ValueError("Expecting a ExprList of coefficients.") if not isinstance(arguments, ExprList): raise ValueError("Expecting a ExprList of arguments.") if not isinstance(coefficient_derivatives, ExprMapping): raise ValueError("Expecting a coefficient-coefficient ExprMapping.") # The coefficient(s) to differentiate w.r.t. and the # argument(s) s.t. D_w[v](e) = d/dtau e(w+tau v)|tau=0 self._w = coefficients.ufl_operands self._v = arguments.ufl_operands self._w2v = {w: v for w, v in zip(self._w, self._v)} # Build more convenient dict {f: df/dw} for each coefficient f # where df/dw is nonzero cd = coefficient_derivatives.ufl_operands self._cd = {cd[2 * i]: cd[2 * i + 1] for i in range(len(cd) // 2)} # Explicitly defining dg/dw == 0 geometric_quantity = GenericDerivativeRuleset.independent_terminal # Explicitly defining da/dw == 0 argument = GenericDerivativeRuleset.independent_terminal
[docs] def coefficient(self, o): """Differentiate a coefficient.""" raise NotImplementedError( "CoordinateDerivative of coefficient in physical space is not implemented." )
[docs] def grad(self, o): """Differentiate a grad.""" raise NotImplementedError("CoordinateDerivative grad in physical space is not implemented.")
[docs] def spatial_coordinate(self, o): """Differentiate a spatial_coordinate.""" do = self._w2v.get(o) # d x /d x => Argument(x.function_space()) if do is not None: return do else: raise NotImplementedError( "CoordinateDerivative found a SpatialCoordinate that is different " "from the one being differentiated." )
[docs] def reference_value(self, o): """Differentiate a reference_value.""" do = self._cd.get(o) if do is not None: return do else: return self.independent_terminal(o)
[docs] def reference_grad(self, g): """Differentiate a reference_grad.""" # d (grad_X(...(x)) / dx => grad_X(...(Argument(x.function_space())) o = g ngrads = 0 while isinstance(o, ReferenceGrad): (o,) = o.ufl_operands ngrads += 1 if not (isinstance(o, SpatialCoordinate) or isinstance(o.ufl_operands[0], FormArgument)): raise ValueError(f"Expecting gradient of a FormArgument, not {ufl_err_str(o)}") def apply_grads(f): for i in range(ngrads): f = ReferenceGrad(f) return f # Find o among all w without any indexing, which makes this # easy for w, v in zip(self._w, self._v): if ( o == w and isinstance(v, ReferenceValue) and isinstance(v.ufl_operands[0], FormArgument) ): # Case: d/dt [w + t v] return apply_grads(v) return self.independent_terminal(o)
[docs] def jacobian(self, o): """Differentiate a jacobian.""" # d (grad_X(x))/d x => grad_X(Argument(x.function_space()) for w, v in zip(self._w, self._v): if extract_unique_domain(o) == extract_unique_domain(w) and isinstance( v.ufl_operands[0], FormArgument ): return ReferenceGrad(v) return self.independent_terminal(o)
[docs]class CoordinateDerivativeRuleDispatcher(MultiFunction): """Dispatcher.""" def __init__(self): """Initialise.""" MultiFunction.__init__(self) self.vcache = defaultdict(dict) self.rcache = defaultdict(dict)
[docs] def terminal(self, o): """Apply to a terminal.""" return o
[docs] def derivative(self, o): """Apply to a derivative.""" raise NotImplementedError(f"Missing derivative handler for {type(o).__name__}.")
expr = MultiFunction.reuse_if_untouched
[docs] def grad(self, o): """Apply to a grad.""" return o
[docs] def reference_grad(self, o): """Apply to a reference_grad.""" return o
[docs] def coefficient_derivative(self, o): """Apply to a coefficient_derivative.""" return o
[docs] def coordinate_derivative(self, o, f, w, v, cd): """Apply to a coordinate_derivative.""" from ufl.algorithms import extract_unique_elements for space in extract_unique_elements(o): if isinstance(space.pullback, CustomPullback): raise NotImplementedError( "CoordinateDerivative is not supported for elements with custom pull back." ) _, w, v, cd = o.ufl_operands rules = CoordinateDerivativeRuleset(w, v, cd) key = (CoordinateDerivativeRuleset, w, v, cd) return map_expr_dag(rules, f, vcache=self.vcache[key], rcache=self.rcache[key])
[docs]def apply_coordinate_derivatives(expression): """Apply coordinate derivatives to an expression.""" rules = CoordinateDerivativeRuleDispatcher() return map_integrand_dags(rules, expression)