"""Implementation of RESCAL."""
from collections.abc import Mapping
from typing import Any, ClassVar
from class_resolver import HintOrType, OptionalKwargs
from torch.nn.init import uniform_
from ..nbase import ERModel
from ...constants import DEFAULT_EMBEDDING_HPO_EMBEDDING_DIM_RANGE
from ...nn import RESCALInteraction
from ...regularizers import LpRegularizer, Regularizer
from ...typing import FloatTensor, Hint, Initializer
__all__ = [
"RESCAL",
]
[docs]
class RESCAL(ERModel[FloatTensor, FloatTensor, FloatTensor]):
r"""An implementation of RESCAL from [nickel2011]_.
RESCAL models entities by $d$-dimensional vectors and relations by $d \times d$-dimensional matrices, both stored
in :class:`~pykeen.nn.representation.Embedding`.
The :class:`~pykeen.nn.modules.RESCALInteraction` function is used to obtain scores from them.
.. note ::
For $E$ entities and $R$ relations, this model requires $Ed + Rd^2$ parameters.
---
citation:
author: Nickel
year: 2011
link: http://www.cip.ifi.lmu.de/~nickel/data/paper-icml2011.pdf
"""
#: The default strategy for optimizing the model's hyper-parameters
hpo_default: ClassVar[Mapping[str, Any]] = dict(
embedding_dim=DEFAULT_EMBEDDING_HPO_EMBEDDING_DIM_RANGE,
)
#: The regularizer used by [nickel2011]_ for for RESCAL
#: According to https://github.com/mnick/rescal.py/blob/master/examples/kinships.py
#: a normalized weight of 10 is used.
regularizer_default: ClassVar[type[Regularizer]] = LpRegularizer
#: The LP settings used by [nickel2011]_ for for RESCAL
regularizer_default_kwargs: ClassVar[Mapping[str, Any]] = dict(
weight=10,
p=2.0,
normalize=True,
)
def __init__(
self,
*,
embedding_dim: int = 50,
entity_initializer: Hint[Initializer] = uniform_,
relation_initializer: Hint[Initializer] = uniform_,
regularizer: HintOrType[Regularizer] = None,
regularizer_kwargs: OptionalKwargs = None,
**kwargs,
) -> None:
r"""Initialize RESCAL.
:param embedding_dim:
the entity embedding dimension $d$. Is usually $d \in [50, 300]$.
:param entity_initializer:
entity initializer function. Defaults to :func:`torch.nn.init.uniform_`
:param relation_initializer:
relation initializer function. Defaults to :func:`torch.nn.init.uniform_`
:param regularizer:
the regularizer. Default to :attr:`pykeen.models.RESCAL.default_regularizer`
:param regularizer_kwargs:
additional keyword-based parameters for the regularizer
:param kwargs:
remaining keyword arguments to forward to :class:`~pykeen.models.ERModel`
.. seealso::
- OpenKE `implementation of RESCAL <https://github.com/thunlp/OpenKE/blob/master/models/RESCAL.py>`_
"""
regularizer = self._instantiate_regularizer(regularizer=regularizer, regularizer_kwargs=regularizer_kwargs)
super().__init__(
interaction=RESCALInteraction,
entity_representations_kwargs=dict(
shape=embedding_dim,
initializer=entity_initializer,
regularizer=regularizer,
),
relation_representations_kwargs=dict(
shape=(embedding_dim, embedding_dim), # d x d matrices
initializer=relation_initializer,
regularizer=regularizer,
),
**kwargs,
)