I'm currently working on my thesis on synthetic data generation. I decided to use google colab to run my model because it faster than my pc.
The model that i'm using are:
CTAB-GAN, https://github.com/Team-TUD/CTAB-GANCTAB-GAN+, https://github.com/Team-TUD/CTAB-GAN-Plus-DP I'm trying to compare both model in the same colab notebook but the problem is that the module I need have the same name and path in the two directory. TheCTABGANclass is in both the model and, except for the name of the first main folder all the others have the same name.
How can I force colab to import a module from a specific path and consequently use both model in the same notebook?
I tried to add the path to both the directory with sys.path.insert: '/content/CTAB-GAN' '/content/CTAB-GAN-Plus-DP'
This operation is done before importing with:
from model.ctabgan import CTABGAN
In this way, colab does not import the second module from CTAB-GAN+ and this means that I cannot use the second model in the notebook or that I can use only one model, depending on which I use first.
So I decided to change the name of the folder and files for the second model. In this way, colab seems to import the correct module from both models. But, when using the model whose folder/files names have been changed, I get this error:DataPrep.__init__() takes 8 positional arguments but 10 were given.
As far as I know, DataPrep is a class used to treat data while they are not ready to be used in training.
