데이터 게이트 검증기와 DAG 연결
This commit is contained in:
@@ -0,0 +1,131 @@
|
||||
"""validate_prediction_accuracy_harness_v2.py — PREDICTION_ACCURACY_HARNESS_VALIDATE_V2
|
||||
|
||||
Temp/prediction_accuracy_harness_v2.json의 기본 구조와 허용된 데이터 게이트 상태를 검증한다.
|
||||
현재는 운영 T+5/T+20 표본이 부족할 수 있으므로 INSUFFICIENT_SAMPLES는 허용한다.
|
||||
"""
|
||||
from __future__ import annotations
|
||||
|
||||
import argparse
|
||||
import json
|
||||
from pathlib import Path
|
||||
from typing import Any
|
||||
|
||||
ROOT = Path(__file__).resolve().parents[1]
|
||||
DEFAULT_INPUT = ROOT / "Temp" / "prediction_accuracy_harness_v2.json"
|
||||
DEFAULT_OUT = ROOT / "Temp" / "validate_prediction_accuracy_harness_v2.json"
|
||||
FORMULA_ID = "PREDICTION_ACCURACY_HARNESS_VALIDATE_V2"
|
||||
ALLOWED_CALIBRATION = {
|
||||
"CALIBRATED",
|
||||
"MONITOR",
|
||||
"PAE_CALIBRATION_REQUIRED",
|
||||
"BUY_PROPOSAL_FROZEN_RECOMMEND",
|
||||
"INSUFFICIENT_SAMPLES",
|
||||
}
|
||||
|
||||
|
||||
def _load(path: Path) -> Any:
|
||||
if not path.exists():
|
||||
return {}
|
||||
try:
|
||||
return json.loads(path.read_text(encoding="utf-8"))
|
||||
except Exception:
|
||||
return {}
|
||||
|
||||
|
||||
def _is_dict(value: Any) -> bool:
|
||||
return isinstance(value, dict)
|
||||
|
||||
|
||||
def _ensure_fields(payload: dict[str, Any], path: str, fields: list[str], errors: list[str]) -> None:
|
||||
block = payload
|
||||
if path:
|
||||
for part in path.split("."):
|
||||
block = block.get(part) if isinstance(block, dict) else None
|
||||
if not isinstance(block, dict):
|
||||
errors.append(f"{path or 'root'} must be object")
|
||||
return
|
||||
for field in fields:
|
||||
if field not in block:
|
||||
errors.append(f"missing field: {path + '.' if path else ''}{field}")
|
||||
|
||||
|
||||
def main() -> int:
|
||||
ap = argparse.ArgumentParser()
|
||||
ap.add_argument("--input", default=str(DEFAULT_INPUT))
|
||||
ap.add_argument("--out", default=str(DEFAULT_OUT))
|
||||
args = ap.parse_args()
|
||||
|
||||
input_path = Path(args.input)
|
||||
input_path = input_path if input_path.is_absolute() else ROOT / input_path
|
||||
out_path = Path(args.out)
|
||||
out_path = out_path if out_path.is_absolute() else ROOT / out_path
|
||||
|
||||
payload = _load(input_path)
|
||||
errors: list[str] = []
|
||||
|
||||
if not _is_dict(payload):
|
||||
errors.append("payload must be object")
|
||||
else:
|
||||
if payload.get("formula_id") != "PREDICTION_ACCURACY_HARNESS_V2":
|
||||
errors.append("formula_id mismatch")
|
||||
|
||||
calibration_state = str(payload.get("calibration_state") or "")
|
||||
if calibration_state not in ALLOWED_CALIBRATION:
|
||||
errors.append(f"calibration_state={calibration_state}")
|
||||
|
||||
for key in [
|
||||
"as_of_date",
|
||||
"data_origin_audit",
|
||||
"windows",
|
||||
"evaluation_methodology",
|
||||
]:
|
||||
if key not in payload:
|
||||
errors.append(f"missing field: {key}")
|
||||
|
||||
audit = payload.get("data_origin_audit")
|
||||
if isinstance(audit, dict):
|
||||
for key in [
|
||||
"operational_sample_count",
|
||||
"replay_sample_count",
|
||||
"untagged_row_count",
|
||||
"unrealized_outcome_row_count",
|
||||
"replay_in_live_stats",
|
||||
"operational_only_accuracy",
|
||||
]:
|
||||
if key not in audit:
|
||||
errors.append(f"missing field: data_origin_audit.{key}")
|
||||
|
||||
for key in [
|
||||
"t1_op_rate", "t1_sample", "t5_op_rate", "t5_sample",
|
||||
"t20_op_rate", "t20_sample", "t20_replay_rate", "t20_replay_sample",
|
||||
"t20_replay_avg_return_pct", "t20_replay_stdev_return_pct",
|
||||
"window_90d_rate",
|
||||
]:
|
||||
if key not in payload:
|
||||
errors.append(f"missing field: {key}")
|
||||
|
||||
windows = payload.get("windows")
|
||||
if isinstance(windows, dict):
|
||||
_ensure_fields(windows, "t1", ["all", "30d", "7d"], errors)
|
||||
_ensure_fields(windows, "t5", ["all", "active_passive", "30d", "90d"], errors)
|
||||
_ensure_fields(windows, "t20", ["operational", "operational_30d", "replay", "replay_return_dist"], errors)
|
||||
else:
|
||||
errors.append("windows must be object")
|
||||
|
||||
t5_sample = payload.get("t5_sample")
|
||||
if isinstance(t5_sample, int) and t5_sample < 30 and calibration_state != "INSUFFICIENT_SAMPLES":
|
||||
errors.append("t5_sample < 30 requires INSUFFICIENT_SAMPLES")
|
||||
|
||||
result = {
|
||||
"formula_id": FORMULA_ID,
|
||||
"gate": "PASS" if not errors else "FAIL",
|
||||
"checked_file": str(Path(args.input).as_posix()),
|
||||
"errors": errors,
|
||||
}
|
||||
out_path.write_text(json.dumps(result, ensure_ascii=False, indent=2) + "\n", encoding="utf-8")
|
||||
print(json.dumps(result, ensure_ascii=False, indent=2))
|
||||
return 0 if not errors else 1
|
||||
|
||||
|
||||
if __name__ == "__main__":
|
||||
raise SystemExit(main())
|
||||
Reference in New Issue
Block a user