CausalDiscoveryToolbox Question about the running time of CGNN

Hi, I am using the package of CGNN, however, I run the sample codes on laptop and server respectively, it seems take a very long time and there is no any reports. The linux system with python 3.7 and pytorch 1.7.1, and the configuration is as follows

I waited the results for nearly 2 hour, is that short or something wrong?

Feb 27 '22 12:02 lishuang1206

Hello,

CGNN can be very time consuming, especially when you don't have GPUs. It is also scaling in time with the number of nodes in your graph (On^3) ; can you set nruns=1 on CGNN ?

Best, Diviyan

Mar 01 '22 07:03 diviyank

Hello,

CGNN can be very time consuming, especially when you don't have GPUs. It is also scaling in time with the number of nodes in your graph (On^3) ; can you set nruns=1 on CGNN ?

Best, Diviyan

Hello, I get it, thank you for your reply, I will try it again. Besides, thank you for your project again, this is really a good tool, hope it will be better.

Best regards, Shuang Li

Mar 01 '22 08:03 lishuang1206