Skip sequences that are less than minimum sequence length

[ajtritt]

Motivation

Sequences less than minimum length were not getting filtered out, which cascaded into a torch error

Traceback (most recent call last):
  File "/pscratch/sd/a/ajtritt/.conda/envs/gtnet-dev/bin/gtnet", line 8, in <module>
    sys.exit(run())
             ^^^^^
  File "/global/u1/a/ajtritt/projects/exabiome/gtnet.git/src/gtnet/main.py", line 45, in run
    func(sys.argv[2:])
  File "/global/u1/a/ajtritt/projects/exabiome/gtnet.git/src/gtnet/classify.py", line 54, in classify
    output = run_torchscript_inference(args.fastas, model, conf_models, window, step, vocab, seqs=args.seqs,
             ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/global/u1/a/ajtritt/projects/exabiome/gtnet.git/src/gtnet/predict.py", line 123, in run_torchscript_inference
    for file_path, seq_name, seq_len, seq_chunks in reader:
  File "/global/u1/a/ajtritt/projects/exabiome/gtnet.git/src/gtnet/sequence.py", line 150, in readfiles
    raise e
  File "/global/u1/a/ajtritt/projects/exabiome/gtnet.git/src/gtnet/sequence.py", line 147, in readfiles
    batches = encoder.encode(values)
              ^^^^^^^^^^^^^^^^^^^^^^
  File "/global/u1/a/ajtritt/projects/exabiome/gtnet.git/src/gtnet/sequence.py", line 61, in encode
    ret = ret.unfold(1, self.window, self.step)
          ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
RuntimeError: maximum size for tensor at dimension 1 is 896 but size is 1024

How to test the behavior?

gtnet classify bin1.fna

Checklist

• Did you update CHANGELOG.md with your changes?
• Have you checked our Contributing document?
• Have you ensured the PR clearly describes the problem and the solution?
• Is your contribution compliant with our coding style? This can be checked running ruff from the source directory.
• Have you checked to ensure that there aren't other open Pull Requests for the same change?
• Have you included the relevant issue number using "Fix #XXX" notation where XXX is the issue number? By including "Fix #XXX" you allow GitHub to close issue #XXX when the PR is merged.

[codecov]

Codecov Report

Merging #44 (cc739dd) into main (8270e52) will decrease coverage by 0.24%.
The diff coverage is 37.50%.

@@            Coverage Diff             @@
##             main      #44      +/-   ##
==========================================
- Coverage   22.29%   22.05%   -0.24%     
==========================================
  Files           7        7              
  Lines         471      476       +5     
  Branches       59       61       +2     
==========================================
  Hits          105      105              
- Misses        359      362       +3     
- Partials        7        9       +2     

Files Changed	Coverage Δ
src/gtnet/sequence.py	72.05% <37.50%> (-2.76%)	⬇️

📣 We’re building smart automated test selection to slash your CI/CD build times. Learn more