Hello Professor.
I'm so sorry to bother you again. When implementing your code, I noticed that you have two versions, one is github version and the other is the final project version under the JSAC folder. The data used by the former is data_git_version.h5 while the data used by the latter is all_data_ct.h5. The former does not distinguish between flow=1 and flow=2 in the dataset.py file, while the latter makes a distinction.
So which version of the code and data should be used to get the results in your paper? I have conducted separate experiments, and the results are different. In my experiments, it may be that the JSAC version is closer, isn't it?
Hello Professor.
I'm so sorry to bother you again. When implementing your code, I noticed that you have two versions, one is github version and the other is the final project version under the JSAC folder. The data used by the former is data_git_version.h5 while the data used by the latter is all_data_ct.h5. The former does not distinguish between flow=1 and flow=2 in the dataset.py file, while the latter makes a distinction.
So which version of the code and data should be used to get the results in your paper? I have conducted separate experiments, and the results are different. In my experiments, it may be that the JSAC version is closer, isn't it?