{
  "acquisition": {
    "acquired_at": "2026-06-26T08:55:35+00:00",
    "fetched_feature_columns": 33,
    "fetched_rows": 1302,
    "openml_data_id": 930,
    "openml_name": "colleges_usnews",
    "openml_url": "https://www.openml.org/d/930",
    "row_policy": "kept_all",
    "source": "openml",
    "source_rows_after_target_cleanup": 1302,
    "stage_path": "/home/ubuntu/hyper/data/tabular/colleges-usnews.csv",
    "staged_feature_columns": 33,
    "staged_rows": 1302,
    "tool": "models.tools.acquire_real_binary_datasets"
  },
  "counts": {
    "dropped_feature_count": 0,
    "n_features": 33,
    "original_columns": 34,
    "original_rows": 1302,
    "rows_after_target_cleanup": 1302,
    "selected_rows": 1302,
    "strata_count": 64,
    "test_rows": 302,
    "train_rows": 1000
  },
  "created_at": "2026-06-26T08:55:35+00:00",
  "dataset_hash": "c98cb891d794dfcd",
  "dropped_feature_columns": [],
  "feature_columns": [
    "FICE",
    "State",
    "Public/private_indicator",
    "Average_Math_SAT_score",
    "Average_Verbal_SAT_score",
    "Average_Combined_SAT_score",
    "Average_ACT_score",
    "First_quartile-Math_SAT",
    "Third_quartile-Math_SAT",
    "First_quartile-Verbal_SAT",
    "Third_quartile-Verbal_SAT",
    "First_quartile-ACT",
    "Third_quartile-ACT",
    "Number_of_applications_received",
    "Number_of_applicants_accepted",
    "Number_of_new_students_enrolled",
    "Pct._new_students_from_top_10Perc_of_H.S._class",
    "Pct._new_students_from_top_25Perc_of_H.S._class",
    "Number_of_fulltime_undergraduates",
    "Number_of_parttime_undergraduates",
    "In-state_tuition",
    "Out-of-state_tuition",
    "Room_and_board_costs",
    "Room_costs",
    "Board_costs",
    "Additional_fees",
    "Estimated_book_costs",
    "Estimated_personal_spending",
    "Pct._of_faculty_with_Ph.D.s",
    "Pct._of_faculty_with_terminal_degree",
    "Student/faculty_ratio",
    "Pct.alumni_who_donate",
    "Instructional_expenditure_per_student"
  ],
  "feature_encodings": {
    "Additional_fees": {
      "impute_value": 264.5,
      "kind": "numeric"
    },
    "Average_ACT_score": {
      "impute_value": 22.0,
      "kind": "numeric"
    },
    "Average_Combined_SAT_score": {
      "impute_value": 957.0,
      "kind": "numeric"
    },
    "Average_Math_SAT_score": {
      "impute_value": 500.0,
      "kind": "numeric"
    },
    "Average_Verbal_SAT_score": {
      "impute_value": 457.0,
      "kind": "numeric"
    },
    "Board_costs": {
      "impute_value": 1980.0,
      "kind": "numeric"
    },
    "Estimated_book_costs": {
      "impute_value": 502.0,
      "kind": "numeric"
    },
    "Estimated_personal_spending": {
      "impute_value": 1250.0,
      "kind": "numeric"
    },
    "FICE": {
      "impute_value": 2649.5,
      "kind": "numeric"
    },
    "First_quartile-ACT": {
      "impute_value": 19.0,
      "kind": "numeric"
    },
    "First_quartile-Math_SAT": {
      "impute_value": 453.0,
      "kind": "numeric"
    },
    "First_quartile-Verbal_SAT": {
      "impute_value": 410.0,
      "kind": "numeric"
    },
    "In-state_tuition": {
      "impute_value": 8050.0,
      "kind": "numeric"
    },
    "Instructional_expenditure_per_student": {
      "impute_value": 7729.0,
      "kind": "numeric"
    },
    "Number_of_applicants_accepted": {
      "impute_value": 1095.0,
      "kind": "numeric"
    },
    "Number_of_applications_received": {
      "impute_value": 1470.0,
      "kind": "numeric"
    },
    "Number_of_fulltime_undergraduates": {
      "impute_value": 1812.0,
      "kind": "numeric"
    },
    "Number_of_new_students_enrolled": {
      "impute_value": 447.0,
      "kind": "numeric"
    },
    "Number_of_parttime_undergraduates": {
      "impute_value": 472.0,
      "kind": "numeric"
    },
    "Out-of-state_tuition": {
      "impute_value": 8670.0,
      "kind": "numeric"
    },
    "Pct._new_students_from_top_10Perc_of_H.S._class": {
      "impute_value": 21.0,
      "kind": "numeric"
    },
    "Pct._new_students_from_top_25Perc_of_H.S._class": {
      "impute_value": 50.0,
      "kind": "numeric"
    },
    "Pct._of_faculty_with_Ph.D.s": {
      "impute_value": 71.0,
      "kind": "numeric"
    },
    "Pct._of_faculty_with_terminal_degree": {
      "impute_value": 77.0,
      "kind": "numeric"
    },
    "Pct.alumni_who_donate": {
      "impute_value": 19.0,
      "kind": "numeric"
    },
    "Public/private_indicator": {
      "impute_value": 2.0,
      "kind": "numeric"
    },
    "Room_and_board_costs": {
      "impute_value": 4030.5,
      "kind": "numeric"
    },
    "Room_costs": {
      "impute_value": 2200.0,
      "kind": "numeric"
    },
    "State": {
      "impute_value": 27.0,
      "kind": "numeric"
    },
    "Student/faculty_ratio": {
      "impute_value": 143.0,
      "kind": "numeric"
    },
    "Third_quartile-ACT": {
      "impute_value": 25.0,
      "kind": "numeric"
    },
    "Third_quartile-Math_SAT": {
      "impute_value": 580.0,
      "kind": "numeric"
    },
    "Third_quartile-Verbal_SAT": {
      "impute_value": 530.0,
      "kind": "numeric"
    }
  },
  "feature_scaling": {
    "constant_columns": [],
    "max": [
      30431.0,
      50.0,
      2.0,
      750.0,
      665.0,
      1410.0,
      31.0,
      740.0,
      785.0,
      630.0,
      720.0,
      29.0,
      35.0,
      48094.0,
      26330.0,
      7425.0,
      98.0,
      100.0,
      31643.0,
      21836.0,
      25750.0,
      25750.0,
      8700.0,
      7400.0,
      6250.0,
      4374.0,
      2340.0,
      6900.0,
      105.0,
      100.0,
      918.0,
      81.0,
      62469.0
    ],
    "method": "minmax over cleaned source rows before sampling",
    "min": [
      1002.0,
      0.0,
      1.0,
      320.0,
      280.0,
      600.0,
      11.0,
      220.0,
      330.0,
      200.0,
      330.0,
      10.0,
      15.0,
      35.0,
      35.0,
      18.0,
      1.0,
      6.0,
      59.0,
      1.0,
      480.0,
      1044.0,
      1260.0,
      500.0,
      531.0,
      9.0,
      90.0,
      75.0,
      8.0,
      20.0,
      23.0,
      0.0,
      1834.0
    ],
    "range": [
      0.0,
      1.0
    ]
  },
  "hash_algorithm": "sha256:first16",
  "headers_removed_from_csv": true,
  "n_features": 33,
  "origin": "real",
  "original_headers": [
    "FICE",
    "State",
    "Public/private_indicator",
    "Average_Math_SAT_score",
    "Average_Verbal_SAT_score",
    "Average_Combined_SAT_score",
    "Average_ACT_score",
    "First_quartile-Math_SAT",
    "Third_quartile-Math_SAT",
    "First_quartile-Verbal_SAT",
    "Third_quartile-Verbal_SAT",
    "First_quartile-ACT",
    "Third_quartile-ACT",
    "Number_of_applications_received",
    "Number_of_applicants_accepted",
    "Number_of_new_students_enrolled",
    "Pct._new_students_from_top_10Perc_of_H.S._class",
    "Pct._new_students_from_top_25Perc_of_H.S._class",
    "Number_of_fulltime_undergraduates",
    "Number_of_parttime_undergraduates",
    "In-state_tuition",
    "Out-of-state_tuition",
    "Room_and_board_costs",
    "Room_costs",
    "Board_costs",
    "Additional_fees",
    "Estimated_book_costs",
    "Estimated_personal_spending",
    "Pct._of_faculty_with_Ph.D.s",
    "Pct._of_faculty_with_terminal_degree",
    "Student/faculty_ratio",
    "Pct.alumni_who_donate",
    "Instructional_expenditure_per_student",
    "target"
  ],
  "output_headers": [
    "FICE",
    "State",
    "Public/private_indicator",
    "Average_Math_SAT_score",
    "Average_Verbal_SAT_score",
    "Average_Combined_SAT_score",
    "Average_ACT_score",
    "First_quartile-Math_SAT",
    "Third_quartile-Math_SAT",
    "First_quartile-Verbal_SAT",
    "Third_quartile-Verbal_SAT",
    "First_quartile-ACT",
    "Third_quartile-ACT",
    "Number_of_applications_received",
    "Number_of_applicants_accepted",
    "Number_of_new_students_enrolled",
    "Pct._new_students_from_top_10Perc_of_H.S._class",
    "Pct._new_students_from_top_25Perc_of_H.S._class",
    "Number_of_fulltime_undergraduates",
    "Number_of_parttime_undergraduates",
    "In-state_tuition",
    "Out-of-state_tuition",
    "Room_and_board_costs",
    "Room_costs",
    "Board_costs",
    "Additional_fees",
    "Estimated_book_costs",
    "Estimated_personal_spending",
    "Pct._of_faculty_with_Ph.D.s",
    "Pct._of_faculty_with_terminal_degree",
    "Student/faculty_ratio",
    "Pct.alumni_who_donate",
    "Instructional_expenditure_per_student",
    "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": 3367608220,
    "source_read_max_rows": null,
    "test_target_rows": 1000,
    "train_target_rows": 1000
  },
  "sha256": "c98cb891d794dfcd11a9634f357b496870d01bb97fe2ee28b83f5aea4b78ea4a",
  "source": "openml",
  "source_file": "colleges-usnews.csv",
  "source_modified_at": "2026-06-26T08:55:35+00:00",
  "source_openml_data_id": 930,
  "source_openml_name": "colleges_usnews",
  "source_path": "/home/ubuntu/hyper/data/tabular/colleges-usnews.csv",
  "source_real_data": true,
  "source_size_bytes": 225678,
  "source_url": "https://www.openml.org/d/930",
  "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": 302,
  "train_rows": 1000
}