misc/py-torch-geometric: New port: Graph neural network library for P…

…yTorch

misc/Makefile

@@ -467,6 +467,7 @@
     SUBDIR += py-tflite
     SUBDIR += py-tflite-support
     SUBDIR += py-toil
+    SUBDIR += py-torch-geometric
     SUBDIR += py-torchvision
     SUBDIR += py-tqdm
     SUBDIR += py-tvm

misc/py-torch-geometric/Makefile

@@ -0,0 +1,32 @@
+PORTNAME=	torch-geometric
+DISTVERSION=	2.3.1
+CATEGORIES=	misc python # machine-learning
+MASTER_SITES=	PYPI
+PKGNAMEPREFIX=	${PYTHON_PKGNAMEPREFIX}
+DISTNAME=	${PORTNAME:S/-/_/}-${PORTVERSION}
+
+MAINTAINER=	yuri@FreeBSD.org
+COMMENT=	Graph neural network library for PyTorch
+WWW=		https://pyg.org/
+
+LICENSE=	MIT
+LICENSE_FILE=	${WRKSRC}/LICENSE
+
+PY_DEPENDS=	${PYTHON_PKGNAMEPREFIX}Jinja2>0:devel/py-Jinja2@${PY_FLAVOR} \
+		${PYNUMPY}  \
+		${PYTHON_PKGNAMEPREFIX}psutil>=5.8.0:sysutils/py-psutil@${PY_FLAVOR} \
+		${PYTHON_PKGNAMEPREFIX}pyparsing>0:devel/py-pyparsing@${PY_FLAVOR} \
+		${PYTHON_PKGNAMEPREFIX}requests>0:www/py-requests@${PY_FLAVOR} \
+		${PYTHON_PKGNAMEPREFIX}scikit-learn>=0:science/py-scikit-learn@${PY_FLAVOR} \
+		${PYTHON_PKGNAMEPREFIX}scipy>0:science/py-scipy@${PY_FLAVOR} \
+		${PYTHON_PKGNAMEPREFIX}tqdm>0:misc/py-tqdm@${PY_FLAVOR}
+BUILD_DEPENDS=	${PY_DEPENDS} \
+		${PYTHON_PKGNAMEPREFIX}wheel>0:devel/py-wheel@${PY_FLAVOR}
+RUN_DEPENDS=	${PY_DEPENDS}
+
+USES=		python:3.7+
+USE_PYTHON=	pep517 autoplist pytest
+
+NO_ARCH=	yes
+
+.include <bsd.port.mk>

misc/py-torch-geometric/distinfo

@@ -0,0 +1,3 @@
+TIMESTAMP = 1685400945
+SHA256 (torch_geometric-2.3.1.tar.gz) = 454fd0bbc128a17a4b9d15010ba9f66d48ec8cd7277991b888a7770263fa125d
+SIZE (torch_geometric-2.3.1.tar.gz) = 661639

misc/py-torch-geometric/pkg-descr

@@ -0,0 +1,12 @@
+PyG (PyTorch Geometric) is a library built upon PyTorch to easily write and
+train Graph Neural Networks (GNNs) for a wide range of applications related
+to structured data.
+
+It consists of various methods for deep learning on graphs and other irregular
+structures, also known as geometric deep learning, from a variety of published
+papers. In addition, it consists of easy-to-use mini-batch loaders for
+operating on many small and single giant graphs, multi GPU-support,
+torch.compile support, DataPipe support, a large number of common benchmark
+datasets (based on simple interfaces to create your own), the GraphGym
+experiment manager, and helpful transforms, both for learning on arbitrary
+graphs as well as on 3D meshes or point clouds.