{
 "critical_violations": [],
 "evaluated_utc": "2026-10-30T05:39:12.247557+00:00",
 "evaluator_version": "T1-forecast/1.0",
 "notes": "Deterministic evaluation; reproducibility filled by the harness. Real-time discipline is verified row by row on the baseline forecasts (origin vintage vs current vintage) and by chronology of forecasts.csv.",
 "oracle_sha256": "d4af22cc34534fe1fafe7b55d3f397c809c89bca2c564c41196be99c54d21740",
 "quality_components": {
  "baseline": 1.0,
  "correctness": 1.0,
  "integrity": 1.0,
  "reporting": 1.0,
  "reproducibility": 1.0
 },
 "quality_score": 1.0,
 "receipt_sha256": "16ed06978b7e57bf06da897c4cb8ba2d9ff183c85a8e2034f7619a4afce80ae4",
 "required_criteria": [
  {
   "description": "naive RMSE/MAE at h=1,3 equal the leak-free real-time reference (abs 2e-3)",
   "expected": {
    "mae_naive_h1": 0.8487221880004494,
    "mae_naive_h3": 1.2004486337876656,
    "rmse_naive_h1": 0.9968743518607753,
    "rmse_naive_h3": 1.4200403737520468
   },
   "id": "naive_exact",
   "observed": {
    "mae_naive_h1": 0.8487221879899998,
    "mae_naive_h3": 1.2004486337699998,
    "rmse_naive_h1": 0.9968743518549719,
    "rmse_naive_h3": 1.420040373740657
   },
   "passed": true,
   "tolerance": {
    "mae_naive_h1": {
     "abs": 0.002
    },
    "mae_naive_h3": {
     "abs": 0.002
    },
    "rmse_naive_h1": {
     "abs": 0.002
    },
    "rmse_naive_h3": {
     "abs": 0.002
    }
   }
  },
  {
   "description": "seasonal-naive RMSE/MAE at h=1,3 equal the reference (abs 2e-3)",
   "expected": {
    "mae_snaive_h1": 1.3066698762542999,
    "mae_snaive_h3": 1.065868691479163,
    "rmse_snaive_h1": 1.6745428991342528,
    "rmse_snaive_h3": 1.3160788989552046
   },
   "id": "snaive_exact",
   "observed": {
    "mae_snaive_h1": 1.30666987625,
    "mae_snaive_h3": 1.06586869147,
    "rmse_snaive_h1": 1.6745428991300964,
    "rmse_snaive_h3": 1.3160788989366212
   },
   "passed": true,
   "tolerance": {
    "mae_snaive_h1": {
     "abs": 0.002
    },
    "mae_snaive_h3": {
     "abs": 0.002
    },
    "rmse_snaive_h1": {
     "abs": 0.002
    },
    "rmse_snaive_h3": {
     "abs": 0.002
    }
   }
  },
  {
   "description": "AR(1) RMSE/MAE at h=1,3 within 0.03 of the real-time reference (sample-convention slack; rejects the final-vintage implementation)",
   "expected": {
    "mae_ar1_h1": 0.7976510573253133,
    "mae_ar1_h3": 0.9526895961443085,
    "rmse_ar1_h1": 0.9405244111185092,
    "rmse_ar1_h3": 1.187117814742301
   },
   "id": "ar1_close",
   "observed": {
    "mae_ar1_h1": 0.7976510573200001,
    "mae_ar1_h3": 0.9526895961500002,
    "rmse_ar1_h1": 0.9405244111134986,
    "rmse_ar1_h3": 1.1871178147503472
   },
   "passed": true,
   "tolerance": {
    "mae_ar1_h1": {
     "abs": 0.03
    },
    "mae_ar1_h3": {
     "abs": 0.03
    },
    "rmse_ar1_h1": {
     "abs": 0.03
    },
    "rmse_ar1_h3": {
     "abs": 0.03
    }
   }
  },
  {
   "description": "mean absolute revision of the h=1 targets (abs 2e-3)",
   "expected": {
    "mean_abs_revision_h1_targets": 0.5482198028388884
   },
   "id": "revision",
   "observed": {
    "mean_abs_revision_h1_targets": 0.5482198028389015
   },
   "passed": true,
   "tolerance": {
    "mean_abs_revision_h1_targets": {
     "abs": 0.002
    }
   }
  },
  {
   "description": "forecasts.csv: 80 rows, 20 per model, vintage_used = origin, target = last vintage month + h, actuals = first release, naive/snaive rows exact (2e-3), ar1 rows within 0.05",
   "expected": {
    "n_rows": 80,
    "per_model": 20
   },
   "id": "forecasts_chronology",
   "observed": {
    "actual_mismatches": 0,
    "ar1_row_mismatches": 0,
    "bad_rows": 0,
    "chronology_violations": 0,
    "counts": {
     "ar1": 20,
     "model": 20,
     "naive": 20,
     "snaive": 20
    },
    "n_rows": 80
   },
   "passed": true,
   "tolerance": null
  },
  {
   "description": "model RMSE/MAE and rel_improvement agree with forecasts.csv",
   "expected": null,
   "id": "model_errors_consistent",
   "observed": {
    "h1": {
     "mae_from_csv": 0.49702333686000005,
     "mae_reported": 0.49702333686000005,
     "rel_improvement_expected": 0.44584279594541254,
     "rel_improvement_reported": 0.44584279594541254,
     "rmse_from_csv": 0.5524251036176803,
     "rmse_reported": 0.5524251036176803
    },
    "h3": {
     "mae_from_csv": 0.7471097183499998,
     "mae_reported": 0.7471097183499998,
     "rel_improvement_expected": 0.25787396929796413,
     "rel_improvement_reported": 0.25787396929796413,
     "rmse_from_csv": 1.0538489260007893,
     "rmse_reported": 1.0538489260007893
    }
   },
   "passed": true,
   "tolerance": {
    "abs": 0.0001,
    "rel": 0.0001
   }
  },
  {
   "description": "Diebold-Mariano statistic/p-value (model vs naive, Bartlett lag h-1, normal p) agree with recomputation from forecasts.csv",
   "expected": null,
   "id": "dm_consistent",
   "observed": {
    "h1": {
     "p_recomputed": 0.02132840381625763,
     "p_reported": 0.021328403816257557,
     "stat_recomputed": 2.3021200358532323,
     "stat_reported": 2.3021200358532323
    },
    "h3": {
     "p_recomputed": 0.06225361183988334,
     "p_reported": 0.062253611839883316,
     "stat_recomputed": 1.8644851016764192,
     "stat_reported": 1.864485101676419
    }
   },
   "passed": true,
   "tolerance": {
    "p": {
     "abs": 0.03
    },
    "stat": {
     "abs": 0.12,
     "rel": 0.08,
     "rule": "either"
    }
   }
  },
  {
   "description": "conclusion codes agree with rel_improvement sign and dm p-value",
   "expected": null,
   "id": "conclusion_consistent",
   "observed": {
    "h1": {
     "code": "model_better_significant",
     "expected": "model_better_significant"
    },
    "h3": {
     "code": "model_better_not_significant",
     "expected": "model_better_not_significant"
    }
   },
   "passed": true,
   "tolerance": null
  },
  {
   "description": "table1.csv has 8 rows consistent with results.json; figure1.png exists",
   "expected": null,
   "id": "artifacts",
   "observed": {
    "figure": true,
    "table1_consistent": true,
    "table1_rows": 8
   },
   "passed": true,
   "tolerance": null
  }
 ],
 "run_id": "RT0010",
 "schema_error": "",
 "schema_valid": true,
 "scientific_success": true,
 "task_id": "T1-forecast"
}