U ʹj=(@s dZddlZddlmZmZdddgZGdddZeZeZe ej e ej dd Z e ej d d Z e ej d d Ze eje eje eje ejddZe eje eje eje ejddZe eje ejddZe eje eje eje ejddZe ejddZe ejddZe ejddZe eje ej ddZ!e ej"e ej"ddZ#e ej$d d!Z%e ej$d"d#Z&e ej'd$d%Z(e ej'd&d'Z)dS)(aF PyTorch provides two global :class:`ConstraintRegistry` objects that link :class:`~torch.distributions.constraints.Constraint` objects to :class:`~torch.distributions.transforms.Transform` objects. These objects both input constraints and return transforms, but they have different guarantees on bijectivity. 1. ``biject_to(constraint)`` looks up a bijective :class:`~torch.distributions.transforms.Transform` from ``constraints.real`` to the given ``constraint``. The returned transform is guaranteed to have ``.bijective = True`` and should implement ``.log_abs_det_jacobian()``. 2. ``transform_to(constraint)`` looks up a not-necessarily bijective :class:`~torch.distributions.transforms.Transform` from ``constraints.real`` to the given ``constraint``. The returned transform is not guaranteed to implement ``.log_abs_det_jacobian()``. The ``transform_to()`` registry is useful for performing unconstrained optimization on constrained parameters of probability distributions, which are indicated by each distribution's ``.arg_constraints`` dict. These transforms often overparameterize a space in order to avoid rotation; they are thus more suitable for coordinate-wise optimization algorithms like Adam:: loc = torch.zeros(100, requires_grad=True) unconstrained = torch.zeros(100, requires_grad=True) scale = transform_to(Normal.arg_constraints['scale'])(unconstrained) loss = -Normal(loc, scale).log_prob(data).sum() The ``biject_to()`` registry is useful for Hamiltonian Monte Carlo, where samples from a probability distribution with constrained ``.support`` are propagated in an unconstrained space, and algorithms are typically rotation invariant.:: dist = Exponential(rate) unconstrained = torch.zeros(100, requires_grad=True) sample = biject_to(dist.support)(unconstrained) potential_energy = -dist.log_prob(sample).sum() .. note:: An example where ``transform_to`` and ``biject_to`` differ is ``constraints.simplex``: ``transform_to(constraints.simplex)`` returns a :class:`~torch.distributions.transforms.SoftmaxTransform` that simply exponentiates and normalizes its inputs; this is a cheap and mostly coordinate-wise operation appropriate for algorithms like SVI. In contrast, ``biject_to(constraints.simplex)`` returns a :class:`~torch.distributions.transforms.StickBreakingTransform` that bijects its input down to a one-fewer-dimensional space; this a more expensive less numerically stable transform but is needed for algorithms like HMC. The ``biject_to`` and ``transform_to`` objects can be extended by user-defined constraints and transforms using their ``.register()`` method either as a function on singleton constraints:: transform_to.register(my_constraint, my_transform) or as a decorator on parameterized constraints:: @transform_to.register(MyConstraintClass) def my_factory(constraint): assert isinstance(constraint, MyConstraintClass) return MyTransform(constraint.param1, constraint.param2) You can create your own registry by creating a new :class:`ConstraintRegistry` object. N) constraints transformsConstraintRegistry biject_to transform_tocs2eZdZdZfddZd ddZddZZS) rz5 Registry to link constraints to transforms. csi|_tdSN) _registrysuper__init__)self __class__f/var/www/html/bwave/python/venv/lib/python3.8/site-packages/torch/distributions/constraint_registry.pyr TszConstraintRegistry.__init__Ncs\|dkrfddSttjr*tttr@ttjsNtd|j<|S)a Registers a :class:`~torch.distributions.constraints.Constraint` subclass in this registry. Usage:: @my_registry.register(MyConstraintClass) def construct_transform(constraint): assert isinstance(constraint, MyConstraint) return MyTransform(constraint.arg_constraints) Args: constraint (subclass of :class:`~torch.distributions.constraints.Constraint`): A subclass of :class:`~torch.distributions.constraints.Constraint`, or a singleton object of the desired class. factory (Callable): A callable that inputs a constraint object and returns a :class:`~torch.distributions.transforms.Transform` object. Ncs |Sr)register)factory constraintr rrkz-ConstraintRegistry.register..zLExpected constraint to be either a Constraint subclass or instance, but got ) isinstancer Constrainttype issubclass TypeErrorrr rrrrrrXs   zConstraintRegistry.registercCsHz|jt|}Wn,tk r>tdt|jddYnX||S)aq Looks up a transform to constrained space, given a constraint object. Usage:: constraint = Normal.arg_constraints['scale'] scale = transform_to(constraint)(torch.zeros(1)) # constrained u = transform_to(constraint).inv(scale) # unconstrained Args: constraint (:class:`~torch.distributions.constraints.Constraint`): A constraint object. Returns: A :class:`~torch.distributions.transforms.Transform` object. Raises: `NotImplementedError` if no transform has been registered. zCannot transform z constraintsN)rrKeyErrorNotImplementedError__name__rrrr__call__{szConstraintRegistry.__call__)N)r __module__ __qualname____doc__r rr __classcell__rrr rrOs  #cCstjSr)rZidentity_transformrrrr_transform_to_realsr%cCst|j}t||jSr)rbase_constraintrIndependentTransformreinterpreted_batch_ndimsrZbase_transformrrr_biject_to_independents  r*cCst|j}t||jSr)rr&rr'r(r)rrr_transform_to_independents  r+cCstSr)r ExpTransformr$rrr_transform_to_positivesr-cCsttt|jdgS)N)rComposeTransformr,AffineTransform lower_boundr$rrr_transform_to_greater_thans  r2cCsttt|jdgS)N)rr/r,r0 upper_boundr$rrr_transform_to_less_thans  r5cCslt|jtjo|jdk}t|jtjo.|jdk}|r@|r@tS|j}|j|j}ttt||gS)Nrr.) rr1numbersNumberr4rZSigmoidTransformr/r0)rZ lower_is_0Z upper_is_1locscalerrr_transform_to_intervals r:cCstSr)rZStickBreakingTransformr$rrr_biject_to_simplexsr;cCstSr)rZSoftmaxTransformr$rrr_transform_to_simplexsr<cCstSr)rZLowerCholeskyTransformr$rrr_transform_to_lower_choleskysr=cCstSr)rZPositiveDefiniteTransformr$rrr_transform_to_positive_definitesr>cCstSr)rZCorrCholeskyTransformr$rrr_transform_to_corr_choleskysr?cCstdd|jD|j|jS)NcSsg|] }t|qSrr.0crrr sz"_biject_to_cat..rZ CatTransformcseqdimlengthsr$rrr_biject_to_cat s rIcCstdd|jD|j|jS)NcSsg|] }t|qSrrrArrrrDsz%_transform_to_cat..rEr$rrr_transform_to_cats rKcCstdd|jD|jS)NcSsg|] }t|qSrr@rArrrrDsz$_biject_to_stack..rZStackTransformrFrGr$rrr_biject_to_stacksrMcCstdd|jD|jS)NcSsg|] }t|qSrrJrArrrrD#sz'_transform_to_stack..rLr$rrr_transform_to_stack srN)*r"r6Ztorch.distributionsrr__all__rrrrrealr%Z independentr*r+ZpositiveZ nonnegativer- greater_thanZgreater_than_eqr2 less_thanr5intervalZhalf_open_intervalr:Zsimplexr;r<Zlower_choleskyr=Zpositive_definiteZpositive_semidefiniter>Z corr_choleskyr?catrIrKstackrMrNrrrrslCI