Dear Olivier,

I am studying your paper on causal generative neural networks, and am attempting implementation of a part of it.

This is regarding the CGNN code in your GitHub repository. I am facing an issue with the code.

I would be very grateful if you could reply and give me a bit of help.

I downloaded your CGNN repository. I followed the instructions, upgraded all packages in requirements.txt.

When I run
python setup.py install develop --user

It gives the following output
running install
running bdist_egg
running egg_info
writing cgnn.egg-info/PKG-INFO
error: [Errno 13] Permission denied: 'cgnn.egg-info/PKG-INFO'

So I installed ran the above command using sudo. The installation then goes through.

But, when I run
sudo python
>>import cgnn

The program terminates without any error.

Does this mean that the install develop step was not properly done?

Could you please help me with this issue?

If it helps,
I am using a virtual environment for python through Anaconda. Python 2.7 and Ubuntu 14.04.5 GNU/Linux 3.13.0

Please let me know if you want any more information from me.

Regards

Hi , dear doc, I want to know if the codes can run on tensorflow with only have CPU, since there are always some errors to imply the unsuccess of the codes, like 'Operation was explicitly assigned to /device:GPU:0 but available devices are [/job:localhost/replica:0/task:0/device:CPU:0]'
thanks!

I find that in CGNN.py on line #292, the reverse_edge function need the parameters of the corresponding nodes of the edge. But nothing is provided in the file.