clefourrier HF staff commited on
Commit
cc85b12
1 Parent(s): 03908b9

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +97 -0
README.md CHANGED
@@ -1,3 +1,100 @@
1
  ---
2
  license: mit
3
  ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
  ---
2
  license: mit
3
  ---
4
+
5
+ # Dataset Card for ogbg-molhiv
6
+
7
+ ## Table of Contents
8
+ - [Table of Contents](#table-of-contents)
9
+ - [Dataset Description](#dataset-description)
10
+ - [Dataset Summary](#dataset-summary)
11
+ - [Supported Tasks and Leaderboards](#supported-tasks-and-leaderboards)
12
+ - [Dataset Structure](#dataset-structure)
13
+ - [Data Properties](#data-properties)
14
+ - [Data Fields](#data-fields)
15
+ - [Data Splits](#data-splits)
16
+ - [Additional Information](#additional-information)
17
+ - [Licensing Information](#licensing-information)
18
+ - [Citation Information](#citation-information)
19
+ - [Contributions](#contributions)
20
+
21
+ ## Dataset Description
22
+
23
+ - **[Homepage](https://ogb.stanford.edu/docs/graphprop/#ogbg-mol)**
24
+ - **[Repository](https://github.com/snap-stanford/ogb):**:
25
+ - **Paper:**: Open Graph Benchmark: Datasets for Machine Learning on Graphs (see citation)
26
+ - **Leaderboard:**: [OGB leaderboard](https://ogb.stanford.edu/docs/leader_graphprop/#ogbg-molhiv) and [Papers with code leaderboard](https://paperswithcode.com/sota/graph-property-prediction-on-ogbg-molhiv)
27
+
28
+ ### Dataset Summary
29
+
30
+ The `ogbg-molhiv` dataset is a small molecular property prediction dataset, adapted from MoleculeNet by teams at Stanford, to be a part of the Open Graph Benchmark.
31
+
32
+ ### Supported Tasks and Leaderboards
33
+
34
+ `ogbg-molhiv` should be used for molecular property prediction (aiming to predict whether molecules inhibit HIV or not), a binary classification task.
35
+ The associated leaderboards are here: [OGB leaderboard](https://ogb.stanford.edu/docs/leader_graphprop/#ogbg-molhiv) and [Papers with code leaderboard](https://paperswithcode.com/sota/graph-property-prediction-on-ogbg-molhiv).
36
+
37
+ ## Dataset Structure
38
+
39
+ ### Data Properties
40
+
41
+ | scale | #graphs | average #nodes | average #edges |
42
+ | small | 41,127 | 25.5 | 27.5 |
43
+ | average node degree | average cluster coefficient | MaxSCC ratio | graph diameter |
44
+ | 2.2 | 0.002 | 0.993 | 12.0 |
45
+
46
+ ### Data Fields
47
+
48
+ Each row of a given file is a graph, with:
49
+ - `x` (list: #nodes x #node-features): nodes
50
+ - `edge_index` (list: 2 x #edges): pairs of nodes constituting edges
51
+ - `edge_attr` (list: #edges x #edge-features): for the aforementioned edges, contains their features
52
+ - `y` (list: 1 x #labels): contains the number of labels available to predict (here 1, equal to zero or one)
53
+ - `num_nodes` (int): number of nodes of the graph
54
+
55
+ ### Data Splits
56
+
57
+ This data comes from the PyGeometric version of the dataset provided by OGB, and follows the provided data splits.
58
+ This information can be found back using
59
+ ```python
60
+ from ogb.graphproppred import PygGraphPropPredDataset
61
+
62
+ dataset = PygGraphPropPredDataset(name = 'ogbg-molhiv')
63
+
64
+ split_idx = dataset.get_idx_split()
65
+ train = dataset[split_idx['train']] # valid, test
66
+ ```
67
+
68
+ ## Additional Information
69
+
70
+ ### Licensing Information
71
+ The dataset has been released under MIT license.
72
+
73
+ ### Citation Information
74
+ ```
75
+ @inproceedings{hu-etal-2020-open,
76
+ author = {Weihua Hu and
77
+ Matthias Fey and
78
+ Marinka Zitnik and
79
+ Yuxiao Dong and
80
+ Hongyu Ren and
81
+ Bowen Liu and
82
+ Michele Catasta and
83
+ Jure Leskovec},
84
+ editor = {Hugo Larochelle and
85
+ Marc Aurelio Ranzato and
86
+ Raia Hadsell and
87
+ Maria{-}Florina Balcan and
88
+ Hsuan{-}Tien Lin},
89
+ title = {Open Graph Benchmark: Datasets for Machine Learning on Graphs},
90
+ booktitle = {Advances in Neural Information Processing Systems 33: Annual Conference
91
+ on Neural Information Processing Systems 2020, NeurIPS 2020, December
92
+ 6-12, 2020, virtual},
93
+ year = {2020},
94
+ url = {https://proceedings.neurips.cc/paper/2020/hash/fb60d411a5c5b72b2e7d3527cfc84fd0-Abstract.html},
95
+ }
96
+ ```
97
+
98
+ ### Contributions
99
+
100
+ Thanks to [@clefourrier](https://github.com/clefourrier) for adding this dataset.