{
  "counts": {
    "dropped_feature_count": 0,
    "n_features": 8,
    "original_columns": 9,
    "original_rows": 155,
    "rows_after_target_cleanup": 155,
    "selected_rows": 155,
    "strata_count": 33,
    "test_rows": 26,
    "train_rows": 129
  },
  "created_at": "2026-06-21T08:32:22+00:00",
  "dataset_hash": "dee1fc8e9afd81c9",
  "dropped_feature_columns": [],
  "feature_columns": [
    "year_zone",
    "year",
    "strip",
    "pdk",
    "damage_rankRJT",
    "damage_rankALL",
    "dry_or_irr",
    "zone"
  ],
  "feature_encodings": {
    "damage_rankALL": {
      "impute_value": 2.0,
      "kind": "numeric"
    },
    "damage_rankRJT": {
      "impute_value": 2.0,
      "kind": "numeric"
    },
    "dry_or_irr": {
      "impute_value": 0.0,
      "kind": "numeric"
    },
    "pdk": {
      "impute_value": 2.0,
      "kind": "numeric"
    },
    "strip": {
      "impute_value": 5.0,
      "kind": "numeric"
    },
    "year": {
      "impute_value": 89.0,
      "kind": "numeric"
    },
    "year_zone": {
      "impute_value": 11.0,
      "kind": "numeric"
    },
    "zone": {
      "impute_value": 1.0,
      "kind": "numeric"
    }
  },
  "feature_scaling": {
    "constant_columns": [],
    "max": [
      20.0,
      92.0,
      10.0,
      5.0,
      5.0,
      5.0,
      2.0,
      2.0
    ],
    "method": "minmax over cleaned source rows before sampling",
    "min": [
      0.0,
      86.0,
      1.0,
      0.0,
      0.0,
      0.0,
      0.0,
      0.0
    ],
    "range": [
      0.0,
      1.0
    ]
  },
  "hash_algorithm": "sha256:first16",
  "headers_removed_from_csv": true,
  "n_features": 8,
  "original_headers": [
    "year_zone",
    "year",
    "strip",
    "pdk",
    "damage_rankRJT",
    "damage_rankALL",
    "dry_or_irr",
    "zone",
    "target"
  ],
  "output_headers": [
    "year_zone",
    "year",
    "strip",
    "pdk",
    "damage_rankRJT",
    "damage_rankALL",
    "dry_or_irr",
    "zone",
    "target"
  ],
  "sampling": {
    "method": "quota stratified by target distribution plus high-variance feature quantile bins",
    "minimum_test_to_train_ratio_when_small": 0.2,
    "seed": 3755268995,
    "test_target_rows": 1000,
    "train_target_rows": 1000
  },
  "sha256": "dee1fc8e9afd81c998bf14180827ac1c52483709d4082bb77b8b8cce7de7b3f8",
  "source_file": "grub-damage.csv",
  "source_modified_at": "2026-06-20T09:30:46+00:00",
  "source_path": "/home/ubuntu/hyper/data/tabular/grub-damage.csv",
  "source_size_bytes": 5929,
  "target_column": "target",
  "target_is_final_column": true,
  "target_names": [
    "target"
  ],
  "target_scaling": {
    "max": 3.0,
    "min": 0.0,
    "mode": "regression",
    "range": [
      0.0,
      1.0
    ]
  },
  "target_source_encoding": {
    "kind": "numeric"
  },
  "task_type": "regression",
  "test_rows": 26,
  "train_rows": 129
}