{
  "per_seed": [
    {
      "seed": 41001,
      "train_prevalence": 0.4462,
      "test_prevalence": 0.4564,
      "baseline_train_prevalence": {
        "brier": 0.24820308000000005,
        "log_loss": 0.6895506486814946,
        "ece_10_equal_width_bins": 0.010199999999999987
      },
      "logistic_candidate": {
        "brier": 0.191956305822283,
        "log_loss": 0.5629905573450384,
        "ece_10_equal_width_bins": 0.015625921644233518
      },
      "permuted_label_negative_control": {
        "brier": 0.2469520813686764,
        "log_loss": 0.6870231820935592,
        "ece_10_equal_width_bins": 0.010682706333177075
      },
      "brier_skill_vs_prevalence": 0.2266159395673778,
      "candidate_beats_baseline_brier": true
    },
    {
      "seed": 41002,
      "train_prevalence": 0.4538,
      "test_prevalence": 0.4462,
      "baseline_train_prevalence": {
        "brier": 0.24716332000000002,
        "log_loss": 0.6874637076353342,
        "ece_10_equal_width_bins": 0.007600000000000218
      },
      "logistic_candidate": {
        "brier": 0.18966789041772467,
        "log_loss": 0.5588335428895983,
        "ece_10_equal_width_bins": 0.013694801617961007
      },
      "permuted_label_negative_control": {
        "brier": 0.24888446780002788,
        "log_loss": 0.6909373006611688,
        "ece_10_equal_width_bins": 0.007664157649948711
      },
      "brier_skill_vs_prevalence": 0.23262120601987124,
      "candidate_beats_baseline_brier": true
    },
    {
      "seed": 41003,
      "train_prevalence": 0.4502,
      "test_prevalence": 0.4614,
      "baseline_train_prevalence": {
        "brier": 0.24863548000000002,
        "log_loss": 0.6904173286960698,
        "ece_10_equal_width_bins": 0.011199999999999988
      },
      "logistic_candidate": {
        "brier": 0.1889263665512363,
        "log_loss": 0.5564883809655683,
        "ece_10_equal_width_bins": 0.015604429972706212
      },
      "permuted_label_negative_control": {
        "brier": 0.24724815430357058,
        "log_loss": 0.6876150747795802,
        "ece_10_equal_width_bins": 0.011199239465623123
      },
      "brier_skill_vs_prevalence": 0.24014719640480808,
      "candidate_beats_baseline_brier": true
    },
    {
      "seed": 41004,
      "train_prevalence": 0.4472,
      "test_prevalence": 0.4522,
      "baseline_train_prevalence": {
        "brier": 0.24774015999999993,
        "log_loss": 0.6886210430756814,
        "ece_10_equal_width_bins": 0.0050000000000000044
      },
      "logistic_candidate": {
        "brier": 0.18493234614465598,
        "log_loss": 0.5472755320854373,
        "ece_10_equal_width_bins": 0.024768693195823872
      },
      "permuted_label_negative_control": {
        "brier": 0.2416695736213931,
        "log_loss": 0.6763317334623488,
        "ece_10_equal_width_bins": 0.007068599694464686
      },
      "brier_skill_vs_prevalence": 0.2535229405492593,
      "candidate_beats_baseline_brier": true
    },
    {
      "seed": 41005,
      "train_prevalence": 0.46,
      "test_prevalence": 0.4548,
      "baseline_train_prevalence": {
        "brier": 0.247984,
        "log_loss": 0.6891099766780086,
        "ece_10_equal_width_bins": 0.0052000000000000934
      },
      "logistic_candidate": {
        "brier": 0.19148930580328807,
        "log_loss": 0.565433349673299,
        "ece_10_equal_width_bins": 0.014511051572677775
      },
      "permuted_label_negative_control": {
        "brier": 0.2505236889033377,
        "log_loss": 0.6942281835661124,
        "ece_10_equal_width_bins": 0.0050095856562485785
      },
      "brier_skill_vs_prevalence": 0.22781588407603692,
      "candidate_beats_baseline_brier": true
    },
    {
      "seed": 41006,
      "train_prevalence": 0.4454,
      "test_prevalence": 0.4624,
      "baseline_train_prevalence": {
        "brier": 0.24887523999999994,
        "log_loss": 0.6909006183533883,
        "ece_10_equal_width_bins": 0.016999999999999904
      },
      "logistic_candidate": {
        "brier": 0.18840251797120278,
        "log_loss": 0.5566780732150203,
        "ece_10_equal_width_bins": 0.02052822478127672
      },
      "permuted_label_negative_control": {
        "brier": 0.2517262770251084,
        "log_loss": 0.69667695857151,
        "ece_10_equal_width_bins": 0.016957876274100003
      },
      "brier_skill_vs_prevalence": 0.24298408322498122,
      "candidate_beats_baseline_brier": true
    },
    {
      "seed": 41007,
      "train_prevalence": 0.466,
      "test_prevalence": 0.4504,
      "baseline_train_prevalence": {
        "brier": 0.24778320000000004,
        "log_loss": 0.688708516279423,
        "ece_10_equal_width_bins": 0.01560000000000017
      },
      "logistic_candidate": {
        "brier": 0.19122486337190664,
        "log_loss": 0.5623658275499777,
        "ece_10_equal_width_bins": 0.019280160492725414
      },
      "permuted_label_negative_control": {
        "brier": 0.2423695495401369,
        "log_loss": 0.6778118500268905,
        "ece_10_equal_width_bins": 0.03446509062089741
      },
      "brier_skill_vs_prevalence": 0.2282573500870656,
      "candidate_beats_baseline_brier": true
    },
    {
      "seed": 41008,
      "train_prevalence": 0.4448,
      "test_prevalence": 0.4506,
      "baseline_train_prevalence": {
        "brier": 0.24759327999999997,
        "log_loss": 0.688326541804596,
        "ece_10_equal_width_bins": 0.005800000000000027
      },
      "logistic_candidate": {
        "brier": 0.19100743850372923,
        "log_loss": 0.5631943350528201,
        "ece_10_equal_width_bins": 0.020905748366551093
      },
      "permuted_label_negative_control": {
        "brier": 0.25186025382642263,
        "log_loss": 0.6969722550123095,
        "ece_10_equal_width_bins": 0.005996308175751142
      },
      "brier_skill_vs_prevalence": 0.2285435270952053,
      "candidate_beats_baseline_brier": true
    },
    {
      "seed": 41009,
      "train_prevalence": 0.4556,
      "test_prevalence": 0.4584,
      "baseline_train_prevalence": {
        "brier": 0.24827728000000002,
        "log_loss": 0.6896978557760698,
        "ece_10_equal_width_bins": 0.002799999999999969
      },
      "logistic_candidate": {
        "brier": 0.1917889315168886,
        "log_loss": 0.5633529377482444,
        "ece_10_equal_width_bins": 0.01206620612597716
      },
      "permuted_label_negative_control": {
        "brier": 0.24819533332762195,
        "log_loss": 0.6895322697999587,
        "ece_10_equal_width_bins": 0.0026306131098430274
      },
      "brier_skill_vs_prevalence": 0.22752121532470226,
      "candidate_beats_baseline_brier": true
    },
    {
      "seed": 41010,
      "train_prevalence": 0.444,
      "test_prevalence": 0.4438,
      "baseline_train_prevalence": {
        "brier": 0.24684160000000002,
        "log_loss": 0.6868170129125417,
        "ece_10_equal_width_bins": 0.0002000000000001445
      },
      "logistic_candidate": {
        "brier": 0.18705066897369338,
        "log_loss": 0.5534459664040332,
        "ece_10_equal_width_bins": 0.008177312604146135
      },
      "permuted_label_negative_control": {
        "brier": 0.2474405342359054,
        "log_loss": 0.6880307375311939,
        "ece_10_equal_width_bins": 0.0019647191085710235
      },
      "brier_skill_vs_prevalence": 0.24222388376313653,
      "candidate_beats_baseline_brier": true
    }
  ],
  "stress_test_per_seed": [
    {
      "seed": 41001,
      "target_shift": "feature mean 0.75, scale 1.25 and intercept -0.8; fully synthetic",
      "shifted_prevalence": 0.4424,
      "complete_model_on_shifted_test": {
        "brier": 0.18658288148547353,
        "log_loss": 0.5529777029291811,
        "ece_10_equal_width_bins": 0.13012356520352666
      },
      "prevalence_baseline_on_shifted_test": {
        "brier": 0.2467136,
        "log_loss": 0.6865605840330763,
        "ece_10_equal_width_bins": 0.005600000000000105
      },
      "complete_model_brier_skill_vs_shifted_baseline": 0.24372680920114043,
      "missingness_rate_per_feature": 0.15,
      "missingness_mechanism": "MCAR; mean imputation uses training-fold feature means only",
      "missing_model_on_shifted_test": {
        "brier": 0.1964883700019899,
        "log_loss": 0.5779679804691268,
        "ece_10_equal_width_bins": 0.10788930628954717
      },
      "missing_model_brier_skill_vs_shifted_baseline": 0.20357706262650332
    },
    {
      "seed": 41002,
      "target_shift": "feature mean 0.75, scale 1.25 and intercept -0.8; fully synthetic",
      "shifted_prevalence": 0.4534,
      "complete_model_on_shifted_test": {
        "brier": 0.18516743514548192,
        "log_loss": 0.5469887506247545,
        "ece_10_equal_width_bins": 0.12063392470626204
      },
      "prevalence_baseline_on_shifted_test": {
        "brier": 0.24782943999999998,
        "log_loss": 0.6887997680709407,
        "ece_10_equal_width_bins": 0.0010000000000000009
      },
      "complete_model_brier_skill_vs_shifted_baseline": 0.2528432653300514,
      "missingness_rate_per_feature": 0.15,
      "missingness_mechanism": "MCAR; mean imputation uses training-fold feature means only",
      "missing_model_on_shifted_test": {
        "brier": 0.19613876440330555,
        "log_loss": 0.5737186057766196,
        "ece_10_equal_width_bins": 0.10667802622302329
      },
      "missing_model_brier_skill_vs_shifted_baseline": 0.2085735883383929
    },
    {
      "seed": 41003,
      "target_shift": "feature mean 0.75, scale 1.25 and intercept -0.8; fully synthetic",
      "shifted_prevalence": 0.4238,
      "complete_model_on_shifted_test": {
        "brier": 0.18980399373359003,
        "log_loss": 0.5597379836981103,
        "ece_10_equal_width_bins": 0.1434070732981428
      },
      "prevalence_baseline_on_shifted_test": {
        "brier": 0.24490112,
        "log_loss": 0.6829239047122283,
        "ece_10_equal_width_bins": 0.026600000000000068
      },
      "complete_model_brier_skill_vs_shifted_baseline": 0.22497702855099222,
      "missingness_rate_per_feature": 0.15,
      "missingness_mechanism": "MCAR; mean imputation uses training-fold feature means only",
      "missing_model_on_shifted_test": {
        "brier": 0.2014362901376758,
        "log_loss": 0.588086958925918,
        "ece_10_equal_width_bins": 0.12227893270733048
      },
      "missing_model_brier_skill_vs_shifted_baseline": 0.17747909794093308
    },
    {
      "seed": 41004,
      "target_shift": "feature mean 0.75, scale 1.25 and intercept -0.8; fully synthetic",
      "shifted_prevalence": 0.4348,
      "complete_model_on_shifted_test": {
        "brier": 0.19454862822132774,
        "log_loss": 0.5724944237420443,
        "ece_10_equal_width_bins": 0.14133451111591483
      },
      "prevalence_baseline_on_shifted_test": {
        "brier": 0.24598612000000003,
        "log_loss": 0.6851009822789933,
        "ece_10_equal_width_bins": 0.01539999999999997
      },
      "complete_model_brier_skill_vs_shifted_baseline": 0.2091072934467696,
      "missingness_rate_per_feature": 0.15,
      "missingness_mechanism": "MCAR; mean imputation uses training-fold feature means only",
      "missing_model_on_shifted_test": {
        "brier": 0.20267484424434384,
        "log_loss": 0.5941879370955491,
        "ece_10_equal_width_bins": 0.11592071197465557
      },
      "missing_model_brier_skill_vs_shifted_baseline": 0.1760720310384024
    },
    {
      "seed": 41005,
      "target_shift": "feature mean 0.75, scale 1.25 and intercept -0.8; fully synthetic",
      "shifted_prevalence": 0.4374,
      "complete_model_on_shifted_test": {
        "brier": 0.1919059150321528,
        "log_loss": 0.5667307176372951,
        "ece_10_equal_width_bins": 0.12972186193602353
      },
      "prevalence_baseline_on_shifted_test": {
        "brier": 0.24609279999999994,
        "log_loss": 0.6853125171921203,
        "ece_10_equal_width_bins": 0.003400000000000014
      },
      "complete_model_brier_skill_vs_shifted_baseline": 0.22018882701097786,
      "missingness_rate_per_feature": 0.15,
      "missingness_mechanism": "MCAR; mean imputation uses training-fold feature means only",
      "missing_model_on_shifted_test": {
        "brier": 0.20429262265421008,
        "log_loss": 0.5961533465882892,
        "ece_10_equal_width_bins": 0.11308812281787198
      },
      "missing_model_brier_skill_vs_shifted_baseline": 0.16985534459273033
    },
    {
      "seed": 41006,
      "target_shift": "feature mean 0.75, scale 1.25 and intercept -0.8; fully synthetic",
      "shifted_prevalence": 0.4428,
      "complete_model_on_shifted_test": {
        "brier": 0.19317635195959723,
        "log_loss": 0.5679917574425007,
        "ece_10_equal_width_bins": 0.13207654354572618
      },
      "prevalence_baseline_on_shifted_test": {
        "brier": 0.24672819999999993,
        "log_loss": 0.6865892331186684,
        "ece_10_equal_width_bins": 0.00019999999999997797
      },
      "complete_model_brier_skill_vs_shifted_baseline": 0.21704794198799615,
      "missingness_rate_per_feature": 0.15,
      "missingness_mechanism": "MCAR; mean imputation uses training-fold feature means only",
      "missing_model_on_shifted_test": {
        "brier": 0.2041410356225677,
        "log_loss": 0.5930042970148818,
        "ece_10_equal_width_bins": 0.11225255934141735
      },
      "missing_model_brier_skill_vs_shifted_baseline": 0.1726076077944566
    },
    {
      "seed": 41007,
      "target_shift": "feature mean 0.75, scale 1.25 and intercept -0.8; fully synthetic",
      "shifted_prevalence": 0.4506,
      "complete_model_on_shifted_test": {
        "brier": 0.18913270312319805,
        "log_loss": 0.5573437385942475,
        "ece_10_equal_width_bins": 0.12978180351898125
      },
      "prevalence_baseline_on_shifted_test": {
        "brier": 0.2475598,
        "log_loss": 0.6882588121056779,
        "ece_10_equal_width_bins": 0.00040000000000001146
      },
      "complete_model_brier_skill_vs_shifted_baseline": 0.23601205396353508,
      "missingness_rate_per_feature": 0.15,
      "missingness_mechanism": "MCAR; mean imputation uses training-fold feature means only",
      "missing_model_on_shifted_test": {
        "brier": 0.19602213967489898,
        "log_loss": 0.5743884843759576,
        "ece_10_equal_width_bins": 0.101298345319253
      },
      "missing_model_brier_skill_vs_shifted_baseline": 0.20818267071269658
    },
    {
      "seed": 41008,
      "target_shift": "feature mean 0.75, scale 1.25 and intercept -0.8; fully synthetic",
      "shifted_prevalence": 0.4282,
      "complete_model_on_shifted_test": {
        "brier": 0.18431883058725138,
        "log_loss": 0.5457976453809533,
        "ece_10_equal_width_bins": 0.11852630147710408
      },
      "prevalence_baseline_on_shifted_test": {
        "brier": 0.24530271999999995,
        "log_loss": 0.6837292806415822,
        "ece_10_equal_width_bins": 0.02140000000000014
      },
      "complete_model_brier_skill_vs_shifted_baseline": 0.24860665798059056,
      "missingness_rate_per_feature": 0.15,
      "missingness_mechanism": "MCAR; mean imputation uses training-fold feature means only",
      "missing_model_on_shifted_test": {
        "brier": 0.19676796859307288,
        "log_loss": 0.5750657635650874,
        "ece_10_equal_width_bins": 0.10362583384397739
      },
      "missing_model_brier_skill_vs_shifted_baseline": 0.1978565562050314
    },
    {
      "seed": 41009,
      "target_shift": "feature mean 0.75, scale 1.25 and intercept -0.8; fully synthetic",
      "shifted_prevalence": 0.4284,
      "complete_model_on_shifted_test": {
        "brier": 0.19579365516961908,
        "log_loss": 0.574937022466066,
        "ece_10_equal_width_bins": 0.15069367793177924
      },
      "prevalence_baseline_on_shifted_test": {
        "brier": 0.24519744000000002,
        "log_loss": 0.6835161185811331,
        "ece_10_equal_width_bins": 0.018000000000000016
      },
      "complete_model_brier_skill_vs_shifted_baseline": 0.20148572852302593,
      "missingness_rate_per_feature": 0.15,
      "missingness_mechanism": "MCAR; mean imputation uses training-fold feature means only",
      "missing_model_on_shifted_test": {
        "brier": 0.2056376131957975,
        "log_loss": 0.5984663517750476,
        "ece_10_equal_width_bins": 0.13262514362554859
      },
      "missing_model_brier_skill_vs_shifted_baseline": 0.16133866162796207
    },
    {
      "seed": 41010,
      "target_shift": "feature mean 0.75, scale 1.25 and intercept -0.8; fully synthetic",
      "shifted_prevalence": 0.4444,
      "complete_model_on_shifted_test": {
        "brier": 0.18948798155598723,
        "log_loss": 0.559595830897851,
        "ece_10_equal_width_bins": 0.12997257799163267
      },
      "prevalence_baseline_on_shifted_test": {
        "brier": 0.24692308,
        "log_loss": 0.6869808642239814,
        "ece_10_equal_width_bins": 0.00379999999999997
      },
      "complete_model_brier_skill_vs_shifted_baseline": 0.2326031995227532,
      "missingness_rate_per_feature": 0.15,
      "missingness_mechanism": "MCAR; mean imputation uses training-fold feature means only",
      "missing_model_on_shifted_test": {
        "brier": 0.20047329397586844,
        "log_loss": 0.5849418644606476,
        "ece_10_equal_width_bins": 0.108683629775771
      },
      "missing_model_brier_skill_vs_shifted_baseline": 0.18811439588446555
    }
  ],
  "original_result_sha256": "28aea3f8eafb63fb10ffe3cc287cc8bde7a15eeb377a5c096c216db95779e410",
  "original_script_sha256": "de26cf35fc9b7930375443cf7b7f3ac87909c83bde95322d0e333e392120c7c3"
}