iMoonLab/DeepHypergraph

Issues

• 大型图的处理

  #40 opened a year ago
  1

• How to transform a igraph or networkx object into hypergraph?

  #29 opened 2 years ago by wanglu2014
  6

• Pip installation error

  #60 opened 6 months ago by dtchang
  4

• How to use batchsize for hypergraph

  #59 opened 7 months ago by A1198482817a
  1

• pip install dhg时出现error：'The 'sklearn' PyPI package is deprecated, use 'scikit-learn' rather than 'sklearn' for pip commands.'

  #55 opened 8 months ago by sds0810
  5

• 如何利用DHG实现整图分类任务

  #30 opened 2 years ago by Li025
  5

• e_weight不满足断言

  #62 opened 5 months ago by unccx
  0

• Has DHG been abandoned?

  #61 opened 6 months ago by dtchang
  0

• Nested hypergraph

  #58 opened 7 months ago by dtchang
  1

• About Directed Hypergraph

  #31 opened 2 years ago by innocentius
  3

• How to get a directed hypergraph

  #47 opened 10 months ago by whyydsforever
  2

• Dataset: Can DHG support loading multi-label node hypergrah datasets?

  #56 opened 8 months ago by zlwu92
  0

• 关于超图子图的接口实现

  #54 opened 8 months ago by talence-zcq
  0

• kindly need help to for several reproduction results

  #24 opened 2 years ago by ShuaiWang97
  3

• 关于UniGCN在计算超边的度的平均值，这是否是一个bug

  #50 opened 8 months ago by HPUtx813
  0

• Can we generate a corresponding hypergraph dataset based on a real-world GNN graph embedding dataset?

  #53 opened 8 months ago by zlwu92
  0

• 您好在使用dhg.models.HGNN出现torch.sparse_coo_tensor的索引超出我的张量尺寸范围，具体的bug如下

  #52 opened 8 months ago by JK-chance
  1

• 超图实现具有不同超边权重的超图卷积模型

  #49 opened 10 months ago by wpx-1
  0

• Didn't concat node's vector with hyperedge's vector before computing the attention of UniGATConv

  #46 opened a year ago by ShangGuanShiHu
  0

• adaptive fusion和coequal fusion的文档问题

  #32 opened 2 years ago by Rechiard
  2

• e_weight=None，agg='softmax' 时Graph类v2v中x=torch.sparse.mm(P,dim=1)应该是x=torch.sparse.softmax(P,dim=1)

  #43 opened a year ago by FuYou0723
  0

• 如何获取训练后节点的向量？

  #42 opened a year ago by jackchenwen
  0

• torch.topk 函数设置 largest=False 是为了找到最小的距离，即最近的邻居。这种方法可能比转换张量为 numpy 数组并使用 scipy.spatial.cKDTree 更有效，因为它避免了 CPU 和 GPU 之间的数据传输。

  #41 opened a year ago
  1

• 中文文档有误

  #39 opened a year ago by unccx
  0

• 我针对你的超图写了个邻接矩阵的接口，想请老哥帮忙看看写的对不对？我自己也做了个些测试，输出的挺对，你介意合并不？

  #38 opened a year ago by 922397935
  2

• 请帮忙看一下画图代码draw_hypergraph（）的参数无法调整如v_line_width和e_line_width，限制输入的格式错了，修改也没变化。

  #37 opened 2 years ago by Longwt123
  2

• CoauthorshipCora dataset label error

  #36 opened 2 years ago by TuDou-PK
  1

• Unable to install dhg using pip

  #35 opened 2 years ago by flash-barry
  2

• mini-batches training for node classification on hypergraph

  #34 opened 2 years ago by ShuaiWang97
  0

• How to use HGNN+ for RL

  #33 opened 2 years ago by FernandezR
  0

• Node classification on multiple hypergraphs

  #20 opened 2 years ago by AbdeslemSmahi
  4

• 超图的顶点度矩阵D_v

  #25 opened 2 years ago by liuzhh1998
  11

• 推荐系统指标部分出现nan，可能是由于除0导致的

  #17 opened 2 years ago by jiishy
  2

• BUG: dhg.Hypergraph.from_feature_kNN() need .detach() for input feature tensor

  #19 opened 2 years ago by realxuyang
  2

• Dataset: Coauthor network from DBLP

  #21 opened 2 years ago by mgao97
  9

• 超图如何采集负样本？

  #28 opened 2 years ago by yw155
  4

• hypergraph.e2v_aggregation_of_group方法出错

  #22 opened 2 years ago by tsubasa0425
  2

• BUG: hg.add_hyperedges can add edges out of vertex index

  #15 opened 2 years ago by realxuyang
  4

• 可视化部分draw_in_poincare_ball(), Norm 运算产生0，导致除法运算报错

  #11 opened 2 years ago by FUNCTIONair-han
  1

• 关于稠密图数据用于训练

  #12 opened 2 years ago by Conor-heming
  13

• Question for the support of dhg for heterogeneous hypergraph

  #9 opened 2 years ago by mgao97
  3

• 中文文档有误

  #18 opened 2 years ago by njtp111
  4

• Can the k in top-k recall (retrieval evaluator) set to a ratio of samples instead of a fixed int?

  #14 opened 2 years ago by realxuyang
  1

• BUG: RetrievalEvaluator.validate_add_batch use the Tensor in CPU

  #13 opened 2 years ago by realxuyang
  1

• I am wondering if the tutorial is still going to be added? It seems like been a year for this repo.

  #2 opened 2 years ago by RussellXie7
  2

• Is this still updated？

  #3 opened 2 years ago by HeLongHuang
  1

• The issue in segmentation

  #4 opened 2 years ago by FeiyangWu6
  4

• Optuna version 1.10 is not exist

  #7 opened 2 years ago by mamintoosi
  1