Spaces:
Running on Zero
Running on Zero
| """Loading precomputed LLR matrices written by scripts/compute_llr_matrices.py.""" | |
| from __future__ import annotations | |
| from pathlib import Path | |
| import numpy as np | |
| import pandas as pd | |
| from src.config import MATRICES_DIR | |
| from src.constants import AMINO_ACIDS | |
| def list_available_accessions(directory: Path = MATRICES_DIR) -> list[str]: | |
| """Accessions with a precomputed `<accession>.npz` in `directory`, sorted.""" | |
| if not directory.exists(): | |
| return [] | |
| return sorted(path.stem for path in directory.glob("*.npz")) | |
| def load_precomputed_matrix(accession: str, directory: Path = MATRICES_DIR) -> tuple[str, pd.DataFrame]: | |
| """Load `<directory>/<accession>.npz`. | |
| Returns `(sequence, matrix)`, where `matrix` has the same shape and row | |
| (`AMINO_ACIDS`) / column (`{residue}{position}`) labeling as | |
| `src.model.calculate_llr_matrix`'s output. | |
| """ | |
| data = np.load(directory / f"{accession}.npz") | |
| sequence = str(data["sequence"]) | |
| columns = [f"{residue}{i + 1}" for i, residue in enumerate(sequence)] | |
| matrix = pd.DataFrame(data["matrix"], index=AMINO_ACIDS, columns=columns) | |
| return sequence, matrix | |