Errors with ICEWS18 scripts from top of repo

[AlexMRuch]

Not sure if I'm missing something, but in RE-Net/data/ICEWS18/get_history_graph.py.py there are two load_quadruples functions that seem to do the same thing.

When I run python data/ICEWS18/get_history_graph.py from the top of the repo directory, I get the following error and traceback:

Traceback (most recent call last):
  File "data/ICEWS18/get_history_graph.py", line 112, in <module>
    train_data, train_times = load_quadruples('', 'train.txt')
  File "data/ICEWS18/get_history_graph.py", line 47, in load_quadruples
    with open(os.path.join(inPath, fileName), 'r') as fr:
FileNotFoundError: [Errno 2] No such file or directory: 'train.txt'

I get the same error with python data/ICEWS18/get_history.py:

Traceback (most recent call last):
  File "data/ICEWS18/get_history.py", line 86, in <module>
    train_data, train_times = load_quadruples('','train.txt')
  File "data/ICEWS18/get_history.py", line 42, in load_quadruples
    with open(os.path.join(inPath, fileName), 'r') as fr:
FileNotFoundError: [Errno 2] No such file or directory: 'train.txt'

Whereas running these scripts from the top of the directory with python data/ICEWS18/get_history.py, etc., does not work, running python get_history.py, etc. from within the data/ICEWS18/ directory is successful. You may want to update your preprocessing notes in https://github.com/INK-USC/re-net#train-and-test to tell people to navigate to the respective data folder first and then run the python script.

Thanks for making this code public!

[AlexMRuch]

Also, when running python train.py -d ICEWS18/ --gpu 1 --model 3 --dropout 0.5 --n-hidden 200 --lr 1e-3 --max-epochs 20 --batch-size 1024 from the top of the repo, I got an FileNotFoundError: [Errno 2] No such file or directory: 'models/ICEWS18/rgcn.pth' error.

Any ideas?

Traceback:

/home/amruch/.local/lib/python3.6/site-packages/sklearn/externals/joblib/externals/cloudpickle/cloudpickle.py:47: DeprecationWarning: the imp module is deprecated in favour of importlib; see the module's documentation for alternative uses
  import imp
Namespace(batch_size=1024, dataset='ICEWS18/', dropout=0.5, gpu=1, grad_norm=1.0, lr=0.001, max_epochs=20, model=3, n_hidden=200, num_k=10, rnn_layers=1, seq_len=10)
start training...
/home/amruch/anaconda3/envs/pytorch/lib/python3.6/site-packages/dgl/base.py:18: UserWarning: Initializer is not set. Use zero initializer instead. To suppress this warning, use `set_initializer` to explicitly specify which initializer to use.
  warnings.warn(msg)
Epoch 0001 | Loss 14.0406 | time 760.5193
/home/amruch/anaconda3/envs/pytorch/lib/python3.6/site-packages/torch/nn/functional.py:1332: UserWarning: nn.functional.sigmoid is deprecated. Use torch.sigmoid instead.
  warnings.warn("nn.functional.sigmoid is deprecated. Use torch.sigmoid instead.")
valid Hits (filtered) @ 1: 0.207555
valid Hits (filtered) @ 3: 0.302348
valid Hits (filtered) @ 10: 0.410349
valid MRR (filtered): 0.276984
valid MR (filtered): 1122.525351
valid Loss: 12.064144
Traceback (most recent call last):
  File "train.py", line 223, in <module>
    train(args)
  File "train.py", line 184, in train
    model_state_file)
  File "/home/amruch/anaconda3/envs/pytorch/lib/python3.6/site-packages/torch/serialization.py", line 218, in save
    return _with_file_like(f, "wb", lambda f: _save(obj, f, pickle_module, pickle_protocol))
  File "/home/amruch/anaconda3/envs/pytorch/lib/python3.6/site-packages/torch/serialization.py", line 141, in _with_file_like
    f = open(f, mode)
FileNotFoundError: [Errno 2] No such file or directory: 'models/ICEWS18/rgcn.pth'

Same happens with the mean approach:

(pytorch) amruch@wit:~/RE-Net$ python train.py -d ICEWS18/ --gpu 1 --model 1 --dropout 0.5 --n-hidden 200 --lr 1e-3 --max-epochs 20 --batch-size 1024
/home/amruch/.local/lib/python3.6/site-packages/sklearn/externals/joblib/externals/cloudpickle/cloudpickle.py:47: DeprecationWarning: the imp module is deprecated in favour of importlib; see the module's documentation for alternative uses
  import imp
Namespace(batch_size=1024, dataset='ICEWS18/', dropout=0.5, gpu=1, grad_norm=1.0, lr=0.001, max_epochs=20, model=1, n_hidden=200, num_k=10, rnn_layers=1, seq_len=10)
start training...
Epoch 0001 | Loss 14.4376 | time 365.4315
/home/amruch/anaconda3/envs/pytorch/lib/python3.6/site-packages/torch/nn/functional.py:1332: UserWarning: nn.functional.sigmoid is deprecated. Use torch.sigmoid instead.
  warnings.warn("nn.functional.sigmoid is deprecated. Use torch.sigmoid instead.")
valid Hits (filtered) @ 1: 0.206316
valid Hits (filtered) @ 3: 0.305294
valid Hits (filtered) @ 10: 0.413893
valid MRR (filtered): 0.277101
valid MR (filtered): 1123.005740
valid Loss: 12.170151
Traceback (most recent call last):
  File "train.py", line 223, in <module>
    train(args)
  File "train.py", line 192, in train
    model_state_file)
  File "/home/amruch/anaconda3/envs/pytorch/lib/python3.6/site-packages/torch/serialization.py", line 218, in save
    return _with_file_like(f, "wb", lambda f: _save(obj, f, pickle_module, pickle_protocol))
  File "/home/amruch/anaconda3/envs/pytorch/lib/python3.6/site-packages/torch/serialization.py", line 141, in _with_file_like
    f = open(f, mode)
FileNotFoundError: [Errno 2] No such file or directory: 'models/ICEWS18/mean.pth'

[AlexMRuch]

For what it's worth, I got the attention model training (and I believe testing too) by manually running mkdir ICEWS18/. Not sure if it would be best to mention one has to do that in the README.md or if you can run an os.mkdir() command in the training script.

[AlexMRuch]

Update: after successfully training and training the attention model from the top of the repo, running the same process for the RGCN aggregator failed during training with the following error:

Traceback (most recent call last):
  File "/home/amruch/RE-Net/test.py", line 147, in <module>
    test(args)
  File "/home/amruch/RE-Net/test.py", line 43, in test
    with open('../data/' + args.dataset+'/test_history_sub.txt', 'rb') as f:
FileNotFoundError: [Errno 2] No such file or directory: '../data/ICEWS18/test_history_sub.txt'

If I navigate to the data/ folder and run the same code (python3 /home/amruch/RE-Net/test.py -d ICEWS18 --gpu 1 --model 3 --n-hidden 200), then it throws this error:

Traceback (most recent call last):
  File "/home/amruch/RE-Net/test.py", line 147, in <module>
    test(args)
  File "/home/amruch/RE-Net/test.py", line 12, in test
    num_nodes, num_rels = utils.get_total_number('./data/' + args.dataset, 'stat.txt')
  File "/home/amruch/RE-Net/utils.py", line 9, in get_total_number
    with open(os.path.join(inPath, fileName), 'r') as fr:
FileNotFoundError: [Errno 2] No such file or directory: './data/ICEWS18/stat.txt'

[AlexMRuch]

I got the script closer to running by editing line ~43 in test.py to read

    if args.model == 3:
        with open('data/' + args.dataset+'/test_history_sub.txt', 'rb') as f:
            s_history_test_data = pickle.load(f)
        with open('data/' + args.dataset+'/test_history_ob.txt', 'rb') as f:
            o_history_test_data = pickle.load(f)

That then gave rise to this error:

start testing:
Using best epoch: 20
/home/amruch/anaconda3/envs/pytorch/lib/python3.6/site-packages/dgl/base.py:18: UserWarning: Initializer is not set. Use zero initializer instead. To suppress this warning, use `set_initializer` to explicitly specify which initializer to use.
  warnings.warn(msg)
/home/amruch/anaconda3/envs/pytorch/lib/python3.6/site-packages/torch/nn/functional.py:1332: UserWarning: nn.functional.sigmoid is deprecated. Use torch.sigmoid instead.
  warnings.warn("nn.functional.sigmoid is deprecated. Use torch.sigmoid instead.")
Traceback (most recent call last):
  File "/home/amruch/RE-Net/test.py", line 147, in <module>
    test(args)
  File "/home/amruch/RE-Net/test.py", line 114, in test
    total_ranks_filter = np.concatenate((total_ranks_filter, ranks_filter))
UnboundLocalError: local variable 'ranks_filter' referenced before assignment

Found out that's because line 114 says anks_filter instead of ranks_filter, which was an easy fix

[AlexMRuch]

Got it running after making those fixes!

[woojeongjin]

Wow thanks for your corrections I updated the repo!

[AlexMRuch]

Glad to help!