{
  "acquisition": {
    "acquired_at": "2026-06-26T09:01:07+00:00",
    "fetched_feature_columns": 36,
    "fetched_rows": 527,
    "openml_data_id": 940,
    "openml_name": "water-treatment",
    "openml_url": "https://www.openml.org/d/940",
    "row_policy": "kept_all",
    "source": "openml",
    "source_rows_after_target_cleanup": 527,
    "stage_path": "/home/ubuntu/hyper/data/tabular/water-treatment.csv",
    "staged_feature_columns": 36,
    "staged_rows": 527,
    "tool": "models.tools.acquire_real_binary_datasets"
  },
  "counts": {
    "dropped_feature_count": 0,
    "n_features": 36,
    "original_columns": 37,
    "original_rows": 527,
    "rows_after_target_cleanup": 527,
    "selected_rows": 527,
    "strata_count": 101,
    "test_rows": 88,
    "train_rows": 439
  },
  "created_at": "2026-06-26T09:01:07+00:00",
  "dataset_hash": "8a13750dca58f373",
  "dropped_feature_columns": [],
  "feature_columns": [
    "ZN-E",
    "PH-E",
    "DBO-E",
    "DQO-E",
    "SS-E",
    "SSV-E",
    "SED-E",
    "COND-E",
    "PH-P",
    "DBO-P",
    "SS-P",
    "SSV-P",
    "SED-P",
    "COND-P",
    "PH-D",
    "DBO-D",
    "DQO-D",
    "SS-D",
    "SSV-D",
    "SED-D",
    "COND-D",
    "PH-S",
    "DBO-S",
    "DQO-S",
    "SS-S",
    "SSV-S",
    "SED-S",
    "COND-S",
    "RD-DBO-P",
    "RD-SS-P",
    "RD-SED-P",
    "RD-DBO-S",
    "RD-DQO-S",
    "RD-DBO-G",
    "RD-DQO-G",
    "RD-SS-G"
  ],
  "feature_encodings": {
    "COND-D": {
      "impute_value": 1428.0,
      "kind": "numeric"
    },
    "COND-E": {
      "impute_value": 1406.0,
      "kind": "numeric"
    },
    "COND-P": {
      "impute_value": 1420.0,
      "kind": "numeric"
    },
    "COND-S": {
      "impute_value": 1432.0,
      "kind": "numeric"
    },
    "DBO-D": {
      "impute_value": 119.0,
      "kind": "numeric"
    },
    "DBO-E": {
      "impute_value": 182.5,
      "kind": "numeric"
    },
    "DBO-P": {
      "impute_value": 197.0,
      "kind": "numeric"
    },
    "DBO-S": {
      "impute_value": 18.0,
      "kind": "numeric"
    },
    "DQO-D": {
      "impute_value": 274.0,
      "kind": "numeric"
    },
    "DQO-E": {
      "impute_value": 400.0,
      "kind": "numeric"
    },
    "DQO-S": {
      "impute_value": 84.0,
      "kind": "numeric"
    },
    "PH-D": {
      "impute_value": 7.8,
      "kind": "numeric"
    },
    "PH-E": {
      "impute_value": 7.8,
      "kind": "numeric"
    },
    "PH-P": {
      "impute_value": 7.8,
      "kind": "numeric"
    },
    "PH-S": {
      "impute_value": 7.7,
      "kind": "numeric"
    },
    "RD-DBO-G": {
      "impute_value": 90.2,
      "kind": "numeric"
    },
    "RD-DBO-P": {
      "impute_value": 39.6,
      "kind": "numeric"
    },
    "RD-DBO-S": {
      "impute_value": 85.4,
      "kind": "numeric"
    },
    "RD-DQO-G": {
      "impute_value": 79.15,
      "kind": "numeric"
    },
    "RD-DQO-S": {
      "impute_value": 69.9,
      "kind": "numeric"
    },
    "RD-SED-P": {
      "impute_value": 93.3,
      "kind": "numeric"
    },
    "RD-SS-G": {
      "impute_value": 90.7,
      "kind": "numeric"
    },
    "RD-SS-P": {
      "impute_value": 59.5,
      "kind": "numeric"
    },
    "SED-D": {
      "impute_value": 0.3,
      "kind": "numeric"
    },
    "SED-E": {
      "impute_value": 4.5,
      "kind": "numeric"
    },
    "SED-P": {
      "impute_value": 4.5,
      "kind": "numeric"
    },
    "SED-S": {
      "impute_value": 0.01,
      "kind": "numeric"
    },
    "SS-D": {
      "impute_value": 90.0,
      "kind": "numeric"
    },
    "SS-E": {
      "impute_value": 196.0,
      "kind": "numeric"
    },
    "SS-P": {
      "impute_value": 220.0,
      "kind": "numeric"
    },
    "SS-S": {
      "impute_value": 19.0,
      "kind": "numeric"
    },
    "SSV-D": {
      "impute_value": 74.5,
      "kind": "numeric"
    },
    "SSV-E": {
      "impute_value": 64.3,
      "kind": "numeric"
    },
    "SSV-P": {
      "impute_value": 62.8,
      "kind": "numeric"
    },
    "SSV-S": {
      "impute_value": 81.15,
      "kind": "numeric"
    },
    "ZN-E": {
      "impute_value": 1.5,
      "kind": "numeric"
    }
  },
  "feature_scaling": {
    "constant_columns": [],
    "max": [
      33.5,
      8.7,
      438.0,
      941.0,
      2008.0,
      85.0,
      36.0,
      3230.0,
      8.5,
      517.0,
      1692.0,
      93.5,
      46.0,
      3170.0,
      8.4,
      285.0,
      511.0,
      244.0,
      100.0,
      3.5,
      3690.0,
      9.7,
      320.0,
      350.0,
      238.0,
      100.0,
      3.5,
      3950.0,
      79.1,
      96.1,
      100.0,
      94.7,
      96.8,
      97.0,
      98.1,
      99.4
    ],
    "method": "minmax over cleaned source rows before sampling",
    "min": [
      0.1,
      6.9,
      31.0,
      81.0,
      98.0,
      13.2,
      0.4,
      651.0,
      7.3,
      32.0,
      104.0,
      7.1,
      1.0,
      646.0,
      7.1,
      26.0,
      80.0,
      49.0,
      20.2,
      0.0,
      85.0,
      7.0,
      3.0,
      9.0,
      6.0,
      29.2,
      0.0,
      683.0,
      0.6,
      5.3,
      7.7,
      8.2,
      1.4,
      19.6,
      19.2,
      10.3
    ],
    "range": [
      0.0,
      1.0
    ]
  },
  "hash_algorithm": "sha256:first16",
  "headers_removed_from_csv": true,
  "n_features": 36,
  "origin": "real",
  "original_headers": [
    "ZN-E",
    "PH-E",
    "DBO-E",
    "DQO-E",
    "SS-E",
    "SSV-E",
    "SED-E",
    "COND-E",
    "PH-P",
    "DBO-P",
    "SS-P",
    "SSV-P",
    "SED-P",
    "COND-P",
    "PH-D",
    "DBO-D",
    "DQO-D",
    "SS-D",
    "SSV-D",
    "SED-D",
    "COND-D",
    "PH-S",
    "DBO-S",
    "DQO-S",
    "SS-S",
    "SSV-S",
    "SED-S",
    "COND-S",
    "RD-DBO-P",
    "RD-SS-P",
    "RD-SED-P",
    "RD-DBO-S",
    "RD-DQO-S",
    "RD-DBO-G",
    "RD-DQO-G",
    "RD-SS-G",
    "target"
  ],
  "output_headers": [
    "ZN-E",
    "PH-E",
    "DBO-E",
    "DQO-E",
    "SS-E",
    "SSV-E",
    "SED-E",
    "COND-E",
    "PH-P",
    "DBO-P",
    "SS-P",
    "SSV-P",
    "SED-P",
    "COND-P",
    "PH-D",
    "DBO-D",
    "DQO-D",
    "SS-D",
    "SSV-D",
    "SED-D",
    "COND-D",
    "PH-S",
    "DBO-S",
    "DQO-S",
    "SS-S",
    "SSV-S",
    "SED-S",
    "COND-S",
    "RD-DBO-P",
    "RD-SS-P",
    "RD-SED-P",
    "RD-DBO-S",
    "RD-DQO-S",
    "RD-DBO-G",
    "RD-DQO-G",
    "RD-SS-G",
    "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": 2334544384,
    "source_read_max_rows": null,
    "test_target_rows": 1000,
    "train_target_rows": 1000
  },
  "sha256": "8a13750dca58f373c2f76ed46264161a3e0712df3f195bbeaa2eb763d8806b5f",
  "source": "openml",
  "source_file": "water-treatment.csv",
  "source_modified_at": "2026-06-26T09:01:07+00:00",
  "source_openml_data_id": 940,
  "source_openml_name": "water-treatment",
  "source_path": "/home/ubuntu/hyper/data/tabular/water-treatment.csv",
  "source_real_data": true,
  "source_size_bytes": 98621,
  "source_url": "https://www.openml.org/d/940",
  "target_column": "target",
  "target_is_final_column": true,
  "target_names": [
    "target"
  ],
  "target_scaling": {
    "class_to_code": {
      "0.0": 0,
      "1.0": 1
    },
    "classes": [
      0.0,
      1.0
    ],
    "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": 88,
  "train_rows": 439
}