{
  "acquisition": {
    "acquired_at": "2026-06-26T09:19:41+00:00",
    "fetched_feature_columns": 32,
    "fetched_rows": 111762,
    "openml_data_id": 44038,
    "openml_name": "road-safety",
    "openml_url": "https://www.openml.org/d/44038",
    "row_policy": "stratified_cap_to_final_max",
    "source": "openml",
    "source_rows_after_target_cleanup": 111762,
    "stage_path": "/home/ubuntu/hyper/data/tabular/road-safety.csv",
    "staged_feature_columns": 32,
    "staged_rows": 2000,
    "tool": "models.tools.acquire_real_binary_datasets"
  },
  "counts": {
    "dropped_feature_count": 0,
    "n_features": 32,
    "original_columns": 33,
    "original_rows": 2000,
    "rows_after_target_cleanup": 2000,
    "selected_rows": 2000,
    "strata_count": 24,
    "test_rows": 1000,
    "train_rows": 1000
  },
  "created_at": "2026-06-26T09:19:41+00:00",
  "dataset_hash": "142ed0a36920369a",
  "dropped_feature_columns": [],
  "feature_columns": [
    "Vehicle_Reference_df_res",
    "Vehicle_Type",
    "Vehicle_Manoeuvre",
    "Vehicle_Location-Restricted_Lane",
    "Hit_Object_in_Carriageway",
    "Hit_Object_off_Carriageway",
    "Was_Vehicle_Left_Hand_Drive?",
    "Age_of_Driver",
    "Age_Band_of_Driver",
    "Engine_Capacity_(CC)",
    "Propulsion_Code",
    "Age_of_Vehicle",
    "Location_Easting_OSGR",
    "Location_Northing_OSGR",
    "Longitude",
    "Latitude",
    "Police_Force",
    "Number_of_Vehicles",
    "Number_of_Casualties",
    "Local_Authority_(District)",
    "1st_Road_Number",
    "2nd_Road_Number",
    "Urban_or_Rural_Area",
    "Vehicle_Reference_df",
    "Casualty_Reference",
    "Sex_of_Casualty",
    "Age_of_Casualty",
    "Age_Band_of_Casualty",
    "Pedestrian_Location",
    "Pedestrian_Movement",
    "Casualty_Type",
    "Casualty_IMD_Decile"
  ],
  "feature_encodings": {
    "1st_Road_Number": {
      "impute_value": 127.0,
      "kind": "numeric"
    },
    "2nd_Road_Number": {
      "impute_value": 0.0,
      "kind": "numeric"
    },
    "Age_Band_of_Casualty": {
      "impute_value": 6.0,
      "kind": "numeric"
    },
    "Age_Band_of_Driver": {
      "impute_value": 7.0,
      "kind": "numeric"
    },
    "Age_of_Casualty": {
      "impute_value": 34.0,
      "kind": "numeric"
    },
    "Age_of_Driver": {
      "impute_value": 38.0,
      "kind": "numeric"
    },
    "Age_of_Vehicle": {
      "impute_value": 8.0,
      "kind": "numeric"
    },
    "Casualty_IMD_Decile": {
      "impute_value": 5.0,
      "kind": "numeric"
    },
    "Casualty_Reference": {
      "impute_value": 1.0,
      "kind": "numeric"
    },
    "Casualty_Type": {
      "impute_value": 9.0,
      "kind": "numeric"
    },
    "Engine_Capacity_(CC)": {
      "impute_value": 1596.0,
      "kind": "numeric"
    },
    "Hit_Object_in_Carriageway": {
      "impute_value": 0.0,
      "kind": "numeric"
    },
    "Hit_Object_off_Carriageway": {
      "impute_value": 0.0,
      "kind": "numeric"
    },
    "Latitude": {
      "impute_value": 52.577420000000004,
      "kind": "numeric"
    },
    "Local_Authority_(District)": {
      "impute_value": 255.0,
      "kind": "numeric"
    },
    "Location_Easting_OSGR": {
      "impute_value": 441451.5,
      "kind": "numeric"
    },
    "Location_Northing_OSGR": {
      "impute_value": 298640.0,
      "kind": "numeric"
    },
    "Longitude": {
      "impute_value": -1.372619,
      "kind": "numeric"
    },
    "Number_of_Casualties": {
      "impute_value": 2.0,
      "kind": "numeric"
    },
    "Number_of_Vehicles": {
      "impute_value": 2.0,
      "kind": "numeric"
    },
    "Pedestrian_Location": {
      "impute_value": 0.0,
      "kind": "numeric"
    },
    "Pedestrian_Movement": {
      "impute_value": 0.0,
      "kind": "numeric"
    },
    "Police_Force": {
      "impute_value": 20.0,
      "kind": "numeric"
    },
    "Propulsion_Code": {
      "impute_value": 1.0,
      "kind": "numeric"
    },
    "Sex_of_Casualty": {
      "impute_value": 1.0,
      "kind": "numeric"
    },
    "Urban_or_Rural_Area": {
      "impute_value": 1.0,
      "kind": "numeric"
    },
    "Vehicle_Location-Restricted_Lane": {
      "impute_value": 0.0,
      "kind": "numeric"
    },
    "Vehicle_Manoeuvre": {
      "impute_value": 16.0,
      "kind": "numeric"
    },
    "Vehicle_Reference_df": {
      "impute_value": 2.0,
      "kind": "numeric"
    },
    "Vehicle_Reference_df_res": {
      "impute_value": 1.0,
      "kind": "numeric"
    },
    "Vehicle_Type": {
      "impute_value": 9.0,
      "kind": "numeric"
    },
    "Was_Vehicle_Left_Hand_Drive?": {
      "impute_value": 1.0,
      "kind": "numeric"
    }
  },
  "feature_scaling": {
    "constant_columns": [],
    "max": [
      33.0,
      97.0,
      18.0,
      9.0,
      12.0,
      11.0,
      2.0,
      92.0,
      11.0,
      16160.0,
      8.0,
      33.0,
      652154.0,
      736978.0,
      1.7223400000000002,
      56.478308,
      98.0,
      37.0,
      36.0,
      932.0,
      9097.0,
      9072.0,
      2.0,
      22.0,
      31.0,
      2.0,
      95.0,
      11.0,
      10.0,
      9.0,
      97.0,
      10.0
    ],
    "method": "minmax over cleaned source rows before sampling",
    "min": [
      1.0,
      2.0,
      1.0,
      0.0,
      0.0,
      0.0,
      1.0,
      16.0,
      4.0,
      49.0,
      1.0,
      1.0,
      150661.0,
      31959.0,
      -5.490548,
      50.13518,
      1.0,
      1.0,
      1.0,
      1.0,
      0.0,
      0.0,
      1.0,
      1.0,
      1.0,
      1.0,
      0.0,
      1.0,
      0.0,
      0.0,
      0.0,
      1.0
    ],
    "range": [
      0.0,
      1.0
    ]
  },
  "hash_algorithm": "sha256:first16",
  "headers_removed_from_csv": true,
  "n_features": 32,
  "origin": "real",
  "original_headers": [
    "Vehicle_Reference_df_res",
    "Vehicle_Type",
    "Vehicle_Manoeuvre",
    "Vehicle_Location-Restricted_Lane",
    "Hit_Object_in_Carriageway",
    "Hit_Object_off_Carriageway",
    "Was_Vehicle_Left_Hand_Drive?",
    "Age_of_Driver",
    "Age_Band_of_Driver",
    "Engine_Capacity_(CC)",
    "Propulsion_Code",
    "Age_of_Vehicle",
    "Location_Easting_OSGR",
    "Location_Northing_OSGR",
    "Longitude",
    "Latitude",
    "Police_Force",
    "Number_of_Vehicles",
    "Number_of_Casualties",
    "Local_Authority_(District)",
    "1st_Road_Number",
    "2nd_Road_Number",
    "Urban_or_Rural_Area",
    "Vehicle_Reference_df",
    "Casualty_Reference",
    "Sex_of_Casualty",
    "Age_of_Casualty",
    "Age_Band_of_Casualty",
    "Pedestrian_Location",
    "Pedestrian_Movement",
    "Casualty_Type",
    "Casualty_IMD_Decile",
    "target"
  ],
  "output_headers": [
    "Vehicle_Reference_df_res",
    "Vehicle_Type",
    "Vehicle_Manoeuvre",
    "Vehicle_Location-Restricted_Lane",
    "Hit_Object_in_Carriageway",
    "Hit_Object_off_Carriageway",
    "Was_Vehicle_Left_Hand_Drive?",
    "Age_of_Driver",
    "Age_Band_of_Driver",
    "Engine_Capacity_(CC)",
    "Propulsion_Code",
    "Age_of_Vehicle",
    "Location_Easting_OSGR",
    "Location_Northing_OSGR",
    "Longitude",
    "Latitude",
    "Police_Force",
    "Number_of_Vehicles",
    "Number_of_Casualties",
    "Local_Authority_(District)",
    "1st_Road_Number",
    "2nd_Road_Number",
    "Urban_or_Rural_Area",
    "Vehicle_Reference_df",
    "Casualty_Reference",
    "Sex_of_Casualty",
    "Age_of_Casualty",
    "Age_Band_of_Casualty",
    "Pedestrian_Location",
    "Pedestrian_Movement",
    "Casualty_Type",
    "Casualty_IMD_Decile",
    "target"
  ],
  "sampling": {
    "final_max_rows": 2000,
    "method": "quota stratified by target distribution plus high-variance feature quantile bins",
    "minimum_final_rows": 101,
    "minimum_test_to_train_ratio_when_small": 0.2,
    "seed": 354154414,
    "source_read_max_rows": null,
    "test_target_rows": 1000,
    "train_target_rows": 1000
  },
  "sha256": "142ed0a36920369ad2411a89a7aa0f77960ed4c020c21a706783aec46d391b29",
  "source": "openml",
  "source_file": "road-safety.csv",
  "source_modified_at": "2026-06-26T09:19:41+00:00",
  "source_openml_data_id": 44038,
  "source_openml_name": "road-safety",
  "source_path": "/home/ubuntu/hyper/data/tabular/road-safety.csv",
  "source_real_data": true,
  "source_size_bytes": 333244,
  "source_url": "https://www.openml.org/d/44038",
  "target_column": "target",
  "target_is_final_column": true,
  "target_names": [
    "target"
  ],
  "target_scaling": {
    "class_to_code": {
      "1.0": 0,
      "2.0": 1
    },
    "classes": [
      1.0,
      2.0
    ],
    "mode": "classification",
    "n_classes": 2,
    "range": [
      0.0,
      1.0
    ],
    "scaled_code_formula": "code / max(1, n_classes - 1)"
  },
  "target_source_encoding": {
    "kind": "numeric"
  },
  "task_type": "binary",
  "test_rows": 1000,
  "train_rows": 1000
}