Refactor HIF

README.md

@@ -8,6 +8,7 @@
 [![project_license][license-shield]][license-url]
 
 [![codecov](https://codecov.io/github/hypernetwork-research-group/hyperbench/graph/badge.svg?token=XE0TB5JMOS)](https://codecov.io/github/hypernetwork-research-group/hyperbench)
+[![CodeFactor](https://www.codefactor.io/repository/github/hypernetwork-research-group/hyperbench/badge)](https://www.codefactor.io/repository/github/hypernetwork-research-group/hyperbench)
 
 ## About the project
 

examples/hgnn.py

@@ -28,7 +28,7 @@
 
     print("Loading and preparing dataset...")
 
-    dataset = AlgebraDataset(sampling_strategy=sampling_strategy, prepare=True)
+    dataset = AlgebraDataset(sampling_strategy=sampling_strategy)
     if verbose:
         print(f"Dataset:\n {dataset.hdata}\n")
 

examples/hyperedge_enricher.py

@@ -5,10 +5,9 @@
 if __name__ == "__main__":
     print("Loading and preparing dataset...\n")
 
-    dataset = AlgebraDataset(sampling_strategy=SamplingStrategy.HYPEREDGE, prepare=True)
+    dataset = AlgebraDataset(sampling_strategy=SamplingStrategy.HYPEREDGE)
 
     print("Enriching hyperedge weights...")
-
     # HyperedgeWeightsEnricher enriches hyperedges with their degree (number of nodes in each hyperedge) as weights.
     # It optionally applies scaling and adds a constant to the weights.
     dataset.enrich_hyperedge_weights(
@@ -25,8 +24,7 @@
 
     print("Enriching hyperedge attributes...")
 
-    # HyperedgeAttrsEnricher adds a feature of 1.0 for each hyperedge,
-    # which can be used as a baseline or for methods that require hyperedge features.
+    # HyperedgeAttrsEnricher adds a feature of 1.0 for each hyperedge, which can be used as a baseline or for methods that require hyperedge features.
     dataset.enrich_hyperedge_attr(
         enricher=HyperedgeAttrsEnricher(),
         enrichment_mode="replace",

examples/hypergcn.py

@@ -28,7 +28,7 @@
 
     print("Loading and preparing dataset...")
 
-    dataset = AlgebraDataset(sampling_strategy=sampling_strategy, prepare=True)
+    dataset = AlgebraDataset(sampling_strategy=sampling_strategy)
     dataset.remove_hyperedges_with_fewer_than_k_nodes(k=2)
     if verbose:
         print(f"Dataset:\n {dataset.hdata}\n")

examples/mlp_common_neighbors.py

@@ -28,7 +28,7 @@
 
     print("Loading and preparing dataset...")
 
-    dataset = AlgebraDataset(sampling_strategy=sampling_strategy, prepare=True)
+    dataset = AlgebraDataset(sampling_strategy=sampling_strategy)
     if verbose:
         print(f"Dataset:\n {dataset.hdata}\n")
 

examples/node2vec.py

@@ -29,7 +29,7 @@
 
     print("Loading and preparing dataset...")
 
-    dataset = AlgebraDataset(sampling_strategy=sampling_strategy, prepare=True)
+    dataset = AlgebraDataset(sampling_strategy=sampling_strategy)
     dataset.remove_hyperedges_with_fewer_than_k_nodes(k=2)
     if verbose:
         print(f"Dataset:\n {dataset.hdata}\n")

examples/node_enricher.py

@@ -5,8 +5,8 @@
 if __name__ == "__main__":
     print("Loading and preparing dataset...")
 
-    dataset = AlgebraDataset(sampling_strategy=SamplingStrategy.HYPEREDGE, prepare=True)
-
+    dataset = AlgebraDataset(sampling_strategy=SamplingStrategy.HYPEREDGE)
+    # NodeEnricher adds features for each node.
     dataset.enrich_node_features(
         enricher=LaplacianPositionalEncodingEnricher(num_features=32),
         enrichment_mode="replace",

hyperbench/data/__init__.py

@@ -1,32 +1,30 @@
-from .dataset import (
-    Dataset,
-    HIFConverter,
-)
+from .dataset import Dataset
+from .hif import HIFLoader, HIFProcessor
 
 from .supported_datasets import (
     AlgebraDataset,
-    AmazonDataset,
-    ContactHighSchoolDataset,
-    ContactPrimarySchoolDataset,
-    CoraDataset,
-    CourseraDataset,
-    DBLPDataset,
-    EmailEnronDataset,
-    EmailW3CDataset,
-    GeometryDataset,
-    GOTDataset,
-    IMDBDataset,
-    MusicBluesReviewsDataset,
-    NBADataset,
-    NDCClassesDataset,
-    NDCSubstancesDataset,
-    PatentDataset,
-    PubmedDataset,
-    RestaurantReviewsDataset,
-    ThreadsAskUbuntuDataset,
-    ThreadsMathsxDataset,
-    TwitterDataset,
-    VegasBarsReviewsDataset,
+    # AmazonDataset,
+    # ContactHighSchoolDataset,
+    # ContactPrimarySchoolDataset,
+    # CoraDataset,
+    # CourseraDataset,
+    # DBLPDataset,
+    # EmailEnronDataset,
+    # EmailW3CDataset,
+    # GeometryDataset,
+    # GOTDataset,
+    # IMDBDataset,
+    # MusicBluesReviewsDataset,
+    # NBADataset,
+    # NDCClassesDataset,
+    # NDCSubstancesDataset,
+    # PatentDataset,
+    # PubmedDataset,
+    # RestaurantReviewsDataset,
+    # ThreadsAskUbuntuDataset,
+    # ThreadsMathsxDataset,
+    # TwitterDataset,
+    # VegasBarsReviewsDataset,
 )
 
 from .loader import DataLoader
@@ -54,7 +52,8 @@
     "EmailW3CDataset",
     "GeometryDataset",
     "GOTDataset",
-    "HIFConverter",
+    "HIFLoader",
+    "HIFProcessor",
     "HyperedgeSampler",
     "IMDBDataset",
     "MusicBluesReviewsDataset",