Train on custom dataset

[mehrazi]

I want to train on a custom dataset for a recommendation task.
The dataset is based on user/group data that users participate in some groups.
The problem is how I should split data for test and train?
when I randomly select some node and edge, I will face this error:
File "/home/BiGI/BiGI_src/utils/GraphMaker.py", line 135, in preprocessing UV_adj = sp.coo_matrix((np.ones(UV_edges.shape[0]), (UV_edges[:, 0], UV_edges[:, 1])), File "/home/.local/lib/python3.8/site-packages/scipy/sparse/coo.py", line 196, in __init__ self._check() File "/home/.local/lib/python3.8/site-packages/scipy/sparse/coo.py", line 283, in _check raise ValueError('row index exceeds matrix dimensions') ValueError: row index exceeds matrix dimensions

[caojiangxia]

Hi~, thank your pay attention to our work.

The reason might be that you should re-index the user/item id from 0, then use the pre-processed datasets as input.

[mehrazi]

Thank you for your reply.
You are right and I'm aware of this problem.
I'm reindexing user/item id from zero, but when I randomly select some user/items for testing, it will be some index removed from training data and cause this problem.
I want to find a way to split user/item ids into two parts for testing and training without this problem happening.
for example, if I have data like this:
1 5
2 6
3 7
4 8
and pick 3 7 for the test, this problem will happen.

[caojiangxia]

You can try this way.

For each user, randomly select his 80% of interactions for training, and others for testing.

If the item id causes the above problem, please repeat the above operation (low probability).

[mehrazi]

The problem is solved.
Thanks for your hints.