Unable to load data

[yangyangdotcom]

I was having the following error when I try to load my data for training

RuntimeError: The 'data' object was created by an older version of PyG. If this error occurred while loading an already existing dataset, remove the 'processed/' directory in the dataset's root folder and try again.

I am new to this , any helps would be much appreciated

Edit:

To make my question clear, the following is my code

Code:

def __getitem__(self, index):
      
      prot_1 = os.path.join(self.processed_dir, self.protein_1[index]+".pt")
      prot_2 = os.path.join(self.processed_dir, self.protein_2[index]+".pt")
      prot_1 = torch.load(glob.glob(prot_1)[0])
      print(prot_1)
      prot_2 = torch.load(glob.glob(prot_2)[0])
      print("stops here")
      return prot_1, prot_2, torch.tensor(self.label[index])

It is at the print(port_1) where the compiler screams error
The self.processed_dir will go to the data folder named "processed" (this is the folder that I created myself) to grab all the '.pt' files

[EdisonLeeeee]

You can simply follow the instruction and "remove the 'processed/' directory in the dataset's root folder and try again".

[yangyangdotcom]

But I do not know where to find the processed/ directory. At first I though it wants me to remove my dataset folder so I delete it even though it doesn't make sense to me. Turns out, it is not my dataset folder that causing this error.

[EdisonLeeeee]

The processed/ directory is located in your dataset's root folder, where root is specified when you initialize the dataset. For example:

dataset = Reddit('data/')

Then your processed dir is located at data/processed

[yangyangdotcom]

Thank you for your reply. So the processed folder is generated by the pyG? I cannot find it in my dataset's root folder

[EdisonLeeeee]

That's weird. Could you show me your code to figure it out?

[yangyangdotcom]

What does the processed/ file actually meant to store? Is it used to store data that is after pre-processing or data that has been feed into the model?

[rusty1s]

raw refers to your raw data in any format, processed refers to the processed data in PyG format, i.e., Data or HeteroData holding PyTorch tensors. It refers to the data after pre-processing/that is inputted into your model.

[yangyangdotcom]

So after I finish the pre-processing phase and got the pre-processed data saved in the processed/ directory, can I use it to as an input to my machine learning model? Or I should move it to other folder?

[rusty1s]

Yes, you can directly make use of it now.

[yangyangdotcom]

I think my problem on this topic is fixed. Thank you all for the help. Much appreciated