{
  "acquisition": {
    "acquired_at": "2026-06-27T23:58:46+00:00",
    "agent_id": "Claude-extra",
    "dedupe_fingerprint": "56abebee8bd7340588ed7e0a5f52542e7f782c28",
    "source_dataset_id": "feynman_test_11",
    "source_dataset_name": "feynman_test_11",
    "source_platform": "pmlb",
    "source_rows": 100000,
    "source_target_column_original": "target",
    "source_url": "https://github.com/EpistasisLab/pmlb/raw/master/datasets/feynman_test_11/feynman_test_11.tsv.gz",
    "staged_rows": 1999,
    "staging_path": "/home/ubuntu/hyper/data/tabular/real_regression_sources/pmlb-feynman-test-11-target.csv",
    "tool": "models.tools.acquire_regression_extra_sources"
  },
  "counts": {
    "dropped_feature_count": 0,
    "n_features": 4,
    "original_columns": 5,
    "original_rows": 1999,
    "rows_after_target_cleanup": 1999,
    "selected_rows": 1999,
    "strata_count": 485,
    "test_rows": 999,
    "train_rows": 1000
  },
  "created_at": "2026-06-27T23:58:46+00:00",
  "dataset_hash": "471f11cc2d38853a",
  "dedupe_fingerprint": "56abebee8bd7340588ed7e0a5f52542e7f782c28",
  "downsampling": {
    "anchors_kept": 10,
    "corr_frob": 0.283188,
    "ks_features_mean": 0.074807,
    "ks_target": 0.012284,
    "method": "hybrid_coreset",
    "minmax_features_preserved": true,
    "minmax_target_preserved": true,
    "n_bins": 50,
    "original_rows": 100000,
    "selected_rows": 1999
  },
  "dropped_feature_columns": [],
  "feature_columns": [
    "I_0",
    "alpha",
    "delta",
    "n"
  ],
  "feature_encodings": {
    "I_0": {
      "impute_value": 2.0600461087833173,
      "kind": "numeric"
    },
    "alpha": {
      "impute_value": 1.950776345909468,
      "kind": "numeric"
    },
    "delta": {
      "impute_value": 1.9604214183938653,
      "kind": "numeric"
    },
    "n": {
      "impute_value": 1.5566134271981684,
      "kind": "numeric"
    }
  },
  "feature_scaling": {
    "constant_columns": [],
    "max": [
      2.999987289599249,
      2.99999664788338,
      2.999991623279082,
      1.999993140429989
    ],
    "method": "minmax over cleaned source rows before sampling",
    "min": [
      1.0000343792921338,
      1.0000062404755132,
      1.00001203405881,
      1.0000005183043128
    ],
    "range": [
      0.0,
      1.0
    ]
  },
  "hash_algorithm": "sha256:first16",
  "headers_removed_from_csv": true,
  "n_features": 4,
  "origin": "real",
  "original_headers": [
    "I_0",
    "alpha",
    "delta",
    "n",
    "target"
  ],
  "output_headers": [
    "I_0",
    "alpha",
    "delta",
    "n",
    "target"
  ],
  "retrieved_at": "2026-06-27T23:54:41+00:00",
  "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": 1177171649,
    "source_read_max_rows": null,
    "test_target_rows": 1000,
    "train_target_rows": 1000
  },
  "sha256": "471f11cc2d38853ae411fbac997c91896e6b22b0e4907c15c6d47c0fc8160fb1",
  "source_dataset_id": "feynman_test_11",
  "source_dataset_name": "feynman_test_11",
  "source_family": "feynman-test",
  "source_file": "pmlb-feynman-test-11-target.csv",
  "source_license": "MIT (PMLB)",
  "source_modified_at": "2026-06-27T23:58:46+00:00",
  "source_path": "/home/ubuntu/hyper/data/tabular/real_regression_sources/pmlb-feynman-test-11-target.csv",
  "source_platform": "pmlb",
  "source_real_data": true,
  "source_resource_id": null,
  "source_rows": 100000,
  "source_size_bytes": 186234,
  "source_target_column_original": "target",
  "source_url": "https://github.com/EpistasisLab/pmlb/raw/master/datasets/feynman_test_11/feynman_test_11.tsv.gz",
  "staged_rows": 1999,
  "staging_path": "/home/ubuntu/hyper/data/tabular/real_regression_sources/pmlb-feynman-test-11-target.csv",
  "target_column": "target",
  "target_generation": "original_numeric_column",
  "target_is_final_column": true,
  "target_names": [
    "target"
  ],
  "target_scaling": {
    "max": 7.926107350151838,
    "min": 0.0149413645546666,
    "mode": "regression",
    "range": [
      0.0,
      1.0
    ]
  },
  "target_source_encoding": {
    "kind": "numeric"
  },
  "task_type": "regression",
  "test_rows": 999,
  "train_rows": 1000
}