"""Datasets from the ILPC 2022 Challenge."""
from docdata import parse_docdata
from .base import UnpackedRemoteDisjointInductiveDataset
__all__ = [
"ILPC2022Small",
"ILPC2022Large",
]
# ZENODO_URL = "https://zenodo.org/record/6321299/files/pykeen/ilpc2022-v1.0.zip"
BASE_URL = "https://raw.githubusercontent.com/pykeen/ilpc2022/master/data"
SMALL_TRAIN_URL = f"{BASE_URL}/small/train.txt"
SMALL_INFERENCE_URL = f"{BASE_URL}/small/inference.txt"
SMALL_INFERENCE_VAL_URL = f"{BASE_URL}/small/inference_validation.txt"
SMALL_INFERENCE_TEST_URL = f"{BASE_URL}/small/inference_test.txt"
LARGE_TRAIN_URL = f"{BASE_URL}/large/train.txt"
LARGE_INFERENCE_URL = f"{BASE_URL}/large/inference.txt"
LARGE_INFERENCE_VAL_URL = f"{BASE_URL}/large/inference_validation.txt"
LARGE_INFERENCE_TEST_URL = f"{BASE_URL}/large/inference_test.txt"
[docs]@parse_docdata
class ILPC2022Small(UnpackedRemoteDisjointInductiveDataset):
"""An inductive link prediction dataset for the ILPC 2022 Challenge.
---
name: ILPC2022 Small
citation:
author: Galkin
year: 2022
link: https://arxiv.org/abs/2203.01520
github: https://github.com/pykeen/ilpc2022
"""
def __init__(self, **kwargs):
"""Initialize the inductive link prediction dataset.
:param kwargs: keyword arguments to forward to the base dataset class, cf. DisjointInductivePathDataset
"""
super().__init__(
transductive_training_url=SMALL_TRAIN_URL,
inductive_inference_url=SMALL_INFERENCE_URL,
inductive_validation_url=SMALL_INFERENCE_VAL_URL,
inductive_testing_url=SMALL_INFERENCE_TEST_URL,
create_inverse_triples=True,
eager=True,
**kwargs,
)
[docs]@parse_docdata
class ILPC2022Large(UnpackedRemoteDisjointInductiveDataset):
"""An inductive link prediction dataset for the ILPC 2022 Challenge.
---
name: ILPC2022 Large
citation:
author: Galkin
year: 2022
link: https://arxiv.org/abs/2203.01520
github: https://github.com/pykeen/ilpc2022
"""
def __init__(self, **kwargs):
"""Initialize the inductive link prediction dataset.
:param kwargs: keyword arguments to forward to the base dataset class, cf. DisjointInductivePathDataset
"""
super().__init__(
transductive_training_url=LARGE_TRAIN_URL,
inductive_inference_url=LARGE_INFERENCE_URL,
inductive_validation_url=LARGE_INFERENCE_VAL_URL,
inductive_testing_url=LARGE_INFERENCE_TEST_URL,
create_inverse_triples=True,
eager=True,
**kwargs,
)
def _main():
for cls in ILPC2022Small, ILPC2022Large:
dataset = cls()
dataset.summarize()
if __name__ == "__main__":
_main()