Datasets:
File size: 4,228 Bytes
84a242a | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 | """Portable paths and provenance safeguards; no scientific computations."""
from __future__ import annotations
import argparse
import hashlib
import importlib.metadata
import json
import platform
import sys
from pathlib import Path
HERE = Path(__file__).resolve().parent
REPOSITORY = HERE.parents[1]
def add_path_arguments(parser: argparse.ArgumentParser, *, inputs: bool = True) -> None:
if inputs:
parser.add_argument("--data-dir", type=Path, default=REPOSITORY / "data/raw")
parser.add_argument("--features", type=Path, default=REPOSITORY / "data/features/complex_features_ligand_protein_d256_seed0.npy")
parser.add_argument("--feature-ids", type=Path, default=REPOSITORY / "data/features/complex_feature_ids.csv",
help="Required one-column CSV of producer-verified IDs in feature-row order.")
parser.add_argument("--descriptors", type=Path, default=REPOSITORY / "data/descriptors/moad_physical_descriptors_filtered.csv")
parser.add_argument("--fold-split", "--fold-file", dest="fold_split", type=Path, default=REPOSITORY / "data/splits/moad_5fold_split_seed42.csv")
parser.add_argument("--output-dir", type=Path, default=REPOSITORY / "outputs/acquisition")
def fingerprint(value: object) -> str:
return hashlib.sha256(json.dumps(value, sort_keys=True, separators=(",", ":"), allow_nan=False).encode()).hexdigest()
def environment_versions() -> dict[str, str]:
versions = {"python": platform.python_version(), "platform": platform.platform()}
for distribution in ("numpy", "pandas", "scikit-learn"):
versions[distribution] = importlib.metadata.version(distribution)
return versions
def previous_checkpoint_size(last_checkpoint: int, checkpoint_targets: dict[int, int]) -> int:
"""Start the next incremental export after the last COMPLETED checkpoint."""
if last_checkpoint == 0:
return 0
if last_checkpoint not in checkpoint_targets:
raise ValueError(f"Unknown completed checkpoint: {last_checkpoint}")
return checkpoint_targets[last_checkpoint]
def verify_existing_run(run_root: Path, expected_fingerprint: str, *, resume: bool) -> bool:
"""Check identity before reading state, skipping completion, or writing outputs.
Return True only for a completed run with matching release provenance.
Historical outputs without this metadata are deliberately not modified.
"""
manifest_path = run_root / "run_state/run_manifest.json"
if not run_root.exists() or not any(run_root.iterdir()):
return False
if not resume:
raise FileExistsError(f"Output run already exists: {run_root}. Use a fresh --output-dir.")
if not manifest_path.exists():
raise RuntimeError(f"Existing run lacks release provenance: {run_root}. Keep historical outputs unchanged and use a fresh --output-dir.")
manifest = json.loads(manifest_path.read_text())
if manifest.get("configuration_fingerprint") != expected_fingerprint:
raise RuntimeError(f"Inputs, code, environment, or run configuration changed: {run_root}. Use a fresh --output-dir.")
marker = run_root / "run_state/complete.json"
if not marker.exists():
return False
complete = json.loads(marker.read_text())
if complete.get("configuration_fingerprint") != expected_fingerprint or complete.get("checkpoints") != 100:
raise RuntimeError(f"Invalid completion marker: {marker}")
final_paths = (
"strict_score_matrices/strict_score_100pct.csv",
"queried_cells/cumulative/queried_100pct.csv",
"queried_cells/newly_queried/queried_100pct.csv",
)
if not all((run_root / path).is_file() for path in final_paths):
raise RuntimeError(f"Completion marker exists but final exports are missing: {run_root}")
return True
def invocation_arguments(args: argparse.Namespace) -> dict[str, object]:
"""Record selected jobs; paths and environment variables are not serialized."""
return {
"methods": list(args.methods), "folds": list(args.folds), "seeds": list(args.seeds),
"smoke_test": bool(getattr(args, "smoke_test", False)),
"resume": not bool(getattr(args, "no_resume", False)),
}
|