torchTS
torchTS copied to clipboard
Time series forecasting with PyTorch
SciPy is only used for some sparse matrix operations in utils/graph.py. These include: - scipy.sparse.coo_matrix - scipy.sparse.csr_matrix - scipy.sparse.diags - scipy.sparse.eye - scipy.sparse.identity (**this just calls eye**) - scipy.sparse.issparse -...
Build tutorials notebooks on how to build DCRNN datasets prior to training it.
This issue is about adding pytest test cases to dcrnn code in the tests/ folder. Some examples include asserting the shape of the input and output matrices of each function...
Distributed training on multiple devices generates this error. ``` dcrnn_gpu.py:16: YAMLLoadWarning: calling yaml.load() without Loader=... is deprecated, as the default Loader is unsafe. Please read https://msg.pyyaml.org/load for full details. config...
Generalize seq2seq class to permit additional recurrent layer types (e.g. GRU). The current implementation has LSTMs hardcoded in both the encoder and decoder.
The current implementation of ```data.py/load_dataset()``` instantiates a standard scaler by default. ```python def load_dataset(dataset_dir, batch_size, val_batch_size=None, test_batch_size=None): if val_batch_size is None: val_batch_size = batch_size if test_batch_size is None: test_batch_size =...