tkipf/pygcn issues and pull requests

#82 - accuracy in the experimental results

Issue - State: open - Opened by Blues-He over 1 year ago - 1 comment

#81 - Cora dataset attributes

Issue - State: open - Opened by mishadr over 1 year ago

#80 - citeseer dataset

Issue - State: open - Opened by zhoulijun11 over 2 years ago

#79 - About the dataset split

Issue - State: open - Opened by DM0815 over 2 years ago - 3 comments

#78 - Why do row normalization instead of column normalization？

Issue - State: open - Opened by Liying1996 almost 3 years ago - 2 comments

#77 - transform to other scope dataset

Issue - State: open - Opened by kimsijin33 about 3 years ago

#76 - Error: 'pybind11' must be installed before running the build.

Issue - State: open - Opened by kimsijin33 about 3 years ago

#75 - Question About fastmode

Issue - State: open - Opened by zhangxingeng over 3 years ago

#74 - Predicting node degree

Issue - State: open - Opened by WilliamKRobert over 3 years ago

#73 - How to do a semi-supervised learning?

Issue - State: open - Opened by cxyccc over 3 years ago - 6 comments

#72 - Invoice node classification / meta-data extraction / single prediction with trained model

Issue - State: open - Opened by Dam2000 over 3 years ago

#71 - In `utils.py` line 36, wouldnt `adj = adj + (adj.T > adj)` also work?

Issue - State: open - Opened by ltbd78 over 3 years ago - 2 comments

#70 - In tensor flow code you used early stopping,isn't it needed in pytorch???

Issue - State: open - Opened by monasolgi almost 4 years ago

#69 - Difference between TF and Pytorch version code

Issue - State: closed - Opened by BonitoW almost 4 years ago - 5 comments

#68 - add renormalization trick option + benchmark notebook

Pull Request - State: open - Opened by mminici about 4 years ago

#67 - citeseer dataset seems doesnot work

Issue - State: open - Opened by welcomeyou2019 about 4 years ago - 2 comments

#66 - question about the adjacency matrix

Issue - State: closed - Opened by Sandydreamsmile about 4 years ago - 2 comments

#65 - Added function to support batch in GCN layer

Pull Request - State: closed - Opened by andreabac3 about 4 years ago

#64 - Normalization of features, batch-wise training, feature extraction

Issue - State: open - Opened by giannispic about 4 years ago

#63 - Hi, does pandas make the data preprocessing more simple?

Issue - State: open - Opened by lonelygoatherd about 4 years ago

#62 - Sort the one-hot-encoding for reproducible results

Pull Request - State: open - Opened by chao1224 over 4 years ago - 3 comments

#61 - Where is Filter parameters in the code?

Issue - State: open - Opened by fansariadeh over 4 years ago - 1 comment

#60 - specifying modes train, validation and test

Issue - State: open - Opened by fansariadeh over 4 years ago - 1 comment

#59 - Values of some parameters

Issue - State: open - Opened by fansariadeh over 4 years ago - 2 comments

#58 - Question about the results

Issue - State: open - Opened by YiBo-stu over 4 years ago - 5 comments

#57 - Using for Regression

Issue - State: open - Opened by Soodeh-K over 4 years ago

#56 - Order of nodes

Issue - State: open - Opened by lenhhoxung86 over 4 years ago - 1 comment

#55 - About symmetrically normalization of adjacency matrix

Issue - State: closed - Opened by empty-id over 4 years ago - 1 comment

#54 - Still curious about the code of building symmetric adjacency matrix

Issue - State: closed - Opened by BilyZ98 almost 5 years ago - 3 comments

#53 - Freeze model performance

Pull Request - State: open - Opened by ChrisZhangcx almost 5 years ago

#52 - Node Classification for Multiple Graphs

Issue - State: closed - Opened by ghost almost 5 years ago - 4 comments

#51 - Why Early Stop is not leverage in this codes?

Issue - State: open - Opened by TranSirius almost 5 years ago

#50 - output = model(features, adj) Are test features involved in the trainning process?

Issue - State: open - Opened by houhouhouhou11 almost 5 years ago - 4 comments

#49 - output = model(features, adj) Are test features involved in the trainning process?

Issue - State: closed - Opened by houhouhouhou11 almost 5 years ago - 1 comment

#48 - The SpecialSpmmFunction Class question

Issue - State: open - Opened by Colorfu1 almost 5 years ago

#47 - What is the difference between two adjacency matrix normalization?

Issue - State: open - Opened by Yfhu1103 almost 5 years ago - 5 comments

#46 - Multiple instance learning and GCN..

Issue - State: open - Opened by manal2222 almost 5 years ago

#45 - What is difference between tensorflow version and pytorch version?

Issue - State: open - Opened by Yangseung about 5 years ago - 1 comment

#44 - question about inputs matrix X, the node feature matirx

Issue - State: open - Opened by PuNeal about 5 years ago - 3 comments

#43 - The intended output

Issue - State: open - Opened by DivyaLekha about 5 years ago

#43 - The intended output

Issue - State: open - Opened by DivyaLekha about 5 years ago

#42 - Xavier initialization use output size instead of input size

Issue - State: open - Opened by rFantic about 5 years ago

#41 - determinism problem

Issue - State: closed - Opened by xuuuluuu over 5 years ago - 3 comments

#40 - GPU usage is 0 when training

Issue - State: closed - Opened by spencerzhang91 over 5 years ago - 1 comment

#39 - Data format

Issue - State: open - Opened by InstantWindy over 5 years ago - 1 comment

#38 - Update train.py

Pull Request - State: closed - Opened by freesinger over 5 years ago

#37 - Question about adj

Issue - State: closed - Opened by zdaiot over 5 years ago - 1 comment

#36 - How to implement residual connections in gcn?

Issue - State: open - Opened by ray075hl over 5 years ago - 2 comments

#35 - How to achieve the training using multi-graph (each graph/mesh from an 3D medical image)?

Issue - State: open - Opened by lileitech over 5 years ago - 1 comment

#34 - Using GCN as encoder alone for multiple small graph data.

Issue - State: open - Opened by ChenChengKuan over 5 years ago - 2 comments

#33 - Allow Using as a Module

Pull Request - State: closed - Opened by Queuecumber over 5 years ago - 1 comment

#32 - No features + sparse labels training

Issue - State: open - Opened by Alekos92 over 5 years ago

#31 - Vocabulary items

Issue - State: open - Opened by gossminn over 5 years ago - 1 comment

#30 - Where is the implementation of the convolved signal matrix

Issue - State: open - Opened by hbsun2113 over 5 years ago - 3 comments

#29 - The cora sites graph is not connected?

Issue - State: closed - Opened by bapriddy over 5 years ago - 6 comments

#28 - I am confused about the dataset

Issue - State: open - Opened by liiiiiiiiil over 5 years ago - 2 comments

#27 - Citeseer data set accuracy

Issue - State: open - Opened by xuhaiyun42 almost 6 years ago - 18 comments

#26 - Could I get the node embedding?

Issue - State: open - Opened by zhiqiangzhongddu almost 6 years ago - 7 comments

#25 - models.py: absolute import layers.GraphConvolution

Pull Request - State: closed - Opened by wbadart almost 6 years ago - 1 comment

#24 - non-square adj matrix

Issue - State: closed - Opened by ahyunSeo almost 6 years ago - 3 comments

#23 - Processing a single graph in batch mode

Issue - State: closed - Opened by motiwari almost 6 years ago - 3 comments

#22 - np negative power

Pull Request - State: closed - Opened by pencoa almost 6 years ago - 5 comments

#21 - load own adjacency matrix and features

Issue - State: open - Opened by xypan1232 almost 6 years ago - 1 comment

#20 - What is different from the original code?

Issue - State: open - Opened by hkthirano almost 6 years ago - 5 comments

#19 - torch.spmm

Issue - State: open - Opened by zhuchangjiang almost 6 years ago - 10 comments

#18 - can I adapt the code to multi-label classificaiton?

Issue - State: open - Opened by xypan1232 almost 6 years ago - 12 comments

#17 - Why test sample number is much more than train / validation sample?

Issue - State: closed - Opened by zhiqiangzhongddu about 6 years ago - 5 comments

#16 - F.log_softmax(x, dim=1) output is not probability?

Issue - State: open - Opened by xypan1232 about 6 years ago - 6 comments

#15 - Small Modifications to make it compatible with python 3.6.3

Pull Request - State: closed - Opened by amar-iastate about 6 years ago - 1 comment

#14 - Scale up to million of nodes

Issue - State: closed - Opened by jthsieh about 6 years ago - 4 comments

#13 - processing with karate datasets

Pull Request - State: open - Opened by ycq091044 about 6 years ago - 5 comments

#12 - datasets partition

Issue - State: open - Opened by 2-4linqian about 6 years ago - 3 comments

#11 - the normalize function in utils.py

Issue - State: open - Opened by wangxiaoyunanne about 6 years ago - 6 comments

#10 - Small changes to make it compatible with PyTorch 0.4.0

Pull Request - State: closed - Opened by nkolot about 6 years ago - 1 comment

#9 - TypeError: expected torch.FloatTensor (got torch.LongTensor)

Issue - State: open - Opened by Cartus over 6 years ago - 3 comments

#8 - Symmetrically normalize adjacency matrix

Issue - State: closed - Opened by XiaoYee over 6 years ago

#7 - How convolutional filters are initialized in spectral domain ?

Issue - State: closed - Opened by pinkfloyd06 over 6 years ago - 1 comment

#6 - PyTorch 0.3 Fix

Pull Request - State: closed - Opened by rusty1s over 6 years ago - 1 comment

#5 - Chebyshev convolution in Pytroch and its kernel size

Issue - State: closed - Opened by pinkfloyd06 over 6 years ago - 2 comments

#4 - Question normalization trick

Issue - State: closed - Opened by Diego999 over 6 years ago - 2 comments

#3 - build symmetric adjacency matrix

Issue - State: closed - Opened by laoreja over 6 years ago - 4 comments

#2 - inconsistent tensor size

Issue - State: closed - Opened by wangxiaoyunanne over 6 years ago - 6 comments

#1 - Batch operations

Issue - State: open - Opened by evanfeinberg almost 7 years ago - 6 comments

GitHub / tkipf/pygcn issues and pull requests