search

muhanzhang / pytorch_DGCNN

PyTorch implementation of DGCNN
MIT License
369 stars 122 forks source link

.mat files #20

Closed LYF14020510036 closed 4 years ago

LYF14020510036 commented 5 years ago

Hello, I am a beginner of graph classification task. Could you tell me where is the code used to generate .mat files in folder data? If my graph benchmark datasets are downloaded from https://ls11-www.cs.tu-dortmund.de/staff/morris/graphkerneldatasets, do I need only the dortmund2txt.m file instead of the mat2txt.m file?

muhanzhang commented 5 years ago

Yes. The .mat files are following the WL kernel toolbox tradition, and are rarely used by graph neural networks. If you use dortmund datasets, you only need "dortmund2txt.m". All graph datasets used here are in txt formats.

LYF14020510036 commented 5 years ago

感谢您的回复，我注意到所用的数据集中原始的 **_graph_labels.txt 中 相同的lable 是在一块的（图1），如果label 是无序的（图2），是不是不支持 image image

muhanzhang commented 5 years ago

并不需要labels有序。dortmund2txt.m 会自动处理成最终的txt格式。

LYF14020510036 commented 5 years ago

感谢您的回复，对我很多帮助，想入坑图分类，节点分类，您的work很赞, 非常感谢开源您的code, 我尝试调试您的代码

LYF14020510036 commented 5 years ago

不好意思，又麻烦你了， 当我使用 dortmund2txt.m 转换数据集 COLLAB 时，报错， 你有遇到过不 image

muhanzhang commented 5 years ago

Hi, please try to "clear all" before processing a new dataset.

LYF14020510036 commented 5 years ago

非常感谢您的指导， 真的解决了