Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Potential Issue with input_length_sequence Implementation in meshnet.data_loader SamplesDataset Class #96

Open
freebob opened this issue Dec 14, 2024 · 0 comments

Comments

@freebob
Copy link

freebob commented Dec 14, 2024

Describe the bug
While reviewing the MeshNet DataLoader, specifically the SamplesDataset class, I came across a section of the code that I believe may not be functioning as intended regarding the input_length_sequence parameter. Here is a brief overview of my observations:

Observed Behavior
The SamplesDataset class is designed to handle sequences of input features with a specified input_length_sequence. However, upon examining the getitem method, it appears that the implementation assumes input_length_sequence=1 regardless of the value provided during initialization. Here is the relevant portion of the code:

def __getitem__(self, idx): ... # Prepare training data. Assume input_sequence_length=1 positions = self._data[trajectory_idx]["pos"][time_idx - 1] # (nnode, dimension) ...
Potential Issue
The comment and subsequent lines in the getitem method suggest that the code is hardcoded to handle only a single time step (input_length_sequence=1). This seems to ignore the input_length_sequence parameter provided during the initialization of the SamplesDataset class. As a result, the functionality intended to support longer input sequences may not be operational.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

1 participant