--- license: cc-by-4.0 task_categories: - other tags: - survival-analysis - causal-inference - treatment-effect-estimation dataset_info: - config_name: actgHC features: - name: setup_key dtype: string - name: scenario dtype: string - name: id dtype: int64 - name: observed_time_month dtype: float64 - name: effect_non_censor dtype: int64 - name: trt dtype: int64 - name: z30 dtype: int64 - name: gender dtype: int64 - name: race dtype: int64 - name: hemo dtype: int64 - name: homo dtype: int64 - name: drugs dtype: int64 - name: str2 dtype: int64 - name: symptom dtype: int64 - name: age dtype: int64 - name: wtkg dtype: float64 - name: karnof dtype: int64 - name: cd40 dtype: int64 - name: cd80 dtype: int64 - name: t0 dtype: float64 - name: t1 dtype: float64 - name: t2 dtype: float64 - name: t3 dtype: float64 - name: t4 dtype: float64 - name: t5 dtype: float64 - name: t6 dtype: float64 - name: t7 dtype: float64 - name: t8 dtype: float64 - name: t9 dtype: float64 - name: e0 dtype: float64 - name: e1 dtype: float64 - name: e2 dtype: float64 - name: e3 dtype: float64 - name: e4 dtype: float64 - name: e5 dtype: float64 - name: e6 dtype: float64 - name: e7 dtype: float64 - name: e8 dtype: float64 - name: e9 dtype: float64 - name: cate_base dtype: float64 - name: summary_json dtype: string splits: - name: train num_bytes: 2210070 num_examples: 3203 download_size: 282257 dataset_size: 2210070 - config_name: actgHC_repeats features: - name: repeat_key dtype: string - name: idx dtype: int64 - name: random_idx0 dtype: int64 - name: random_idx1 dtype: int64 - name: random_idx2 dtype: int64 - name: random_idx3 dtype: int64 - name: random_idx4 dtype: int64 - name: random_idx5 dtype: int64 - name: random_idx6 dtype: int64 - name: random_idx7 dtype: int64 - name: random_idx8 dtype: int64 - name: random_idx9 dtype: int64 splits: - name: train num_bytes: 336315 num_examples: 3203 download_size: 102559 dataset_size: 336315 - config_name: actgHC_splits features: - name: setup_key dtype: string - name: scenario dtype: string - name: id dtype: int64 - name: observed_time_month dtype: float64 - name: effect_non_censor dtype: int64 - name: trt dtype: int64 - name: z30 dtype: int64 - name: gender dtype: int64 - name: race dtype: int64 - name: hemo dtype: int64 - name: homo dtype: int64 - name: drugs dtype: int64 - name: str2 dtype: int64 - name: symptom dtype: int64 - name: age dtype: int64 - name: wtkg dtype: float64 - name: karnof dtype: int64 - name: cd40 dtype: int64 - name: cd80 dtype: int64 - name: t0 dtype: float64 - name: t1 dtype: float64 - name: t2 dtype: float64 - name: t3 dtype: float64 - name: t4 dtype: float64 - name: t5 dtype: float64 - name: t6 dtype: float64 - name: t7 dtype: float64 - name: t8 dtype: float64 - name: t9 dtype: float64 - name: e0 dtype: float64 - name: e1 dtype: float64 - name: e2 dtype: float64 - name: e3 dtype: float64 - name: e4 dtype: float64 - name: e5 dtype: float64 - name: e6 dtype: float64 - name: e7 dtype: float64 - name: e8 dtype: float64 - name: e9 dtype: float64 - name: cate_base dtype: float64 splits: - name: train_0 num_bytes: 535330 num_examples: 1598 - name: val_0 num_bytes: 268000 num_examples: 800 - name: test_0 num_bytes: 268670 num_examples: 802 - name: train_1 num_bytes: 535330 num_examples: 1598 - name: val_1 num_bytes: 268000 num_examples: 800 - name: test_1 num_bytes: 268670 num_examples: 802 - name: train_2 num_bytes: 535330 num_examples: 1598 - name: val_2 num_bytes: 268000 num_examples: 800 - name: test_2 num_bytes: 268670 num_examples: 802 - name: train_3 num_bytes: 536335 num_examples: 1601 - name: val_3 num_bytes: 266995 num_examples: 797 - name: test_3 num_bytes: 268670 num_examples: 802 - name: train_4 num_bytes: 536335 num_examples: 1601 - name: val_4 num_bytes: 268000 num_examples: 800 - name: test_4 num_bytes: 267665 num_examples: 799 - name: train_5 num_bytes: 536335 num_examples: 1601 - name: val_5 num_bytes: 266995 num_examples: 797 - name: test_5 num_bytes: 268670 num_examples: 802 - name: train_6 num_bytes: 536335 num_examples: 1601 - name: val_6 num_bytes: 268000 num_examples: 800 - name: test_6 num_bytes: 267665 num_examples: 799 - name: train_7 num_bytes: 536000 num_examples: 1600 - name: val_7 num_bytes: 268000 num_examples: 800 - name: test_7 num_bytes: 268000 num_examples: 800 - name: train_8 num_bytes: 535330 num_examples: 1598 - name: val_8 num_bytes: 268000 num_examples: 800 - name: test_8 num_bytes: 268670 num_examples: 802 - name: train_9 num_bytes: 536335 num_examples: 1601 - name: val_9 num_bytes: 267330 num_examples: 798 - name: test_9 num_bytes: 268335 num_examples: 801 download_size: 3586688 dataset_size: 10720000 - config_name: actgLC features: - name: setup_key dtype: string - name: scenario dtype: string - name: id dtype: int64 - name: observed_time_month dtype: float64 - name: effect_non_censor dtype: int64 - name: trt dtype: int64 - name: z30 dtype: int64 - name: gender dtype: int64 - name: race dtype: int64 - name: hemo dtype: int64 - name: homo dtype: int64 - name: drugs dtype: int64 - name: str2 dtype: int64 - name: symptom dtype: int64 - name: age dtype: int64 - name: wtkg dtype: float64 - name: karnof dtype: int64 - name: cd40 dtype: int64 - name: cd80 dtype: int64 - name: t0 dtype: float64 - name: t1 dtype: float64 - name: t2 dtype: float64 - name: t3 dtype: float64 - name: t4 dtype: float64 - name: t5 dtype: float64 - name: t6 dtype: float64 - name: t7 dtype: float64 - name: t8 dtype: float64 - name: t9 dtype: float64 - name: e0 dtype: float64 - name: e1 dtype: float64 - name: e2 dtype: float64 - name: e3 dtype: float64 - name: e4 dtype: float64 - name: e5 dtype: float64 - name: e6 dtype: float64 - name: e7 dtype: float64 - name: e8 dtype: float64 - name: e9 dtype: float64 - name: cate_base dtype: float64 - name: summary_json dtype: string splits: - name: train num_bytes: 2210070 num_examples: 3203 download_size: 282257 dataset_size: 2210070 - config_name: actgLC_repeats features: - name: repeat_key dtype: string - name: idx dtype: int64 - name: random_idx0 dtype: int64 - name: random_idx1 dtype: int64 - name: random_idx2 dtype: int64 - name: random_idx3 dtype: int64 - name: random_idx4 dtype: int64 - name: random_idx5 dtype: int64 - name: random_idx6 dtype: int64 - name: random_idx7 dtype: int64 - name: random_idx8 dtype: int64 - name: random_idx9 dtype: int64 splits: - name: train num_bytes: 336315 num_examples: 3203 download_size: 102559 dataset_size: 336315 - config_name: actgLC_splits features: - name: setup_key dtype: string - name: scenario dtype: string - name: id dtype: int64 - name: observed_time_month dtype: float64 - name: effect_non_censor dtype: int64 - name: trt dtype: int64 - name: z30 dtype: int64 - name: gender dtype: int64 - name: race dtype: int64 - name: hemo dtype: int64 - name: homo dtype: int64 - name: drugs dtype: int64 - name: str2 dtype: int64 - name: symptom dtype: int64 - name: age dtype: int64 - name: wtkg dtype: float64 - name: karnof dtype: int64 - name: cd40 dtype: int64 - name: cd80 dtype: int64 - name: t0 dtype: float64 - name: t1 dtype: float64 - name: t2 dtype: float64 - name: t3 dtype: float64 - name: t4 dtype: float64 - name: t5 dtype: float64 - name: t6 dtype: float64 - name: t7 dtype: float64 - name: t8 dtype: float64 - name: t9 dtype: float64 - name: e0 dtype: float64 - name: e1 dtype: float64 - name: e2 dtype: float64 - name: e3 dtype: float64 - name: e4 dtype: float64 - name: e5 dtype: float64 - name: e6 dtype: float64 - name: e7 dtype: float64 - name: e8 dtype: float64 - name: e9 dtype: float64 - name: cate_base dtype: float64 splits: - name: train_0 num_bytes: 535330 num_examples: 1598 - name: val_0 num_bytes: 268000 num_examples: 800 - name: test_0 num_bytes: 268670 num_examples: 802 - name: train_1 num_bytes: 535330 num_examples: 1598 - name: val_1 num_bytes: 268000 num_examples: 800 - name: test_1 num_bytes: 268670 num_examples: 802 - name: train_2 num_bytes: 535330 num_examples: 1598 - name: val_2 num_bytes: 268000 num_examples: 800 - name: test_2 num_bytes: 268670 num_examples: 802 - name: train_3 num_bytes: 536335 num_examples: 1601 - name: val_3 num_bytes: 266995 num_examples: 797 - name: test_3 num_bytes: 268670 num_examples: 802 - name: train_4 num_bytes: 536335 num_examples: 1601 - name: val_4 num_bytes: 268000 num_examples: 800 - name: test_4 num_bytes: 267665 num_examples: 799 - name: train_5 num_bytes: 536335 num_examples: 1601 - name: val_5 num_bytes: 266995 num_examples: 797 - name: test_5 num_bytes: 268670 num_examples: 802 - name: train_6 num_bytes: 536335 num_examples: 1601 - name: val_6 num_bytes: 268000 num_examples: 800 - name: test_6 num_bytes: 267665 num_examples: 799 - name: train_7 num_bytes: 536000 num_examples: 1600 - name: val_7 num_bytes: 268000 num_examples: 800 - name: test_7 num_bytes: 268000 num_examples: 800 - name: train_8 num_bytes: 535330 num_examples: 1598 - name: val_8 num_bytes: 268000 num_examples: 800 - name: test_8 num_bytes: 268670 num_examples: 802 - name: train_9 num_bytes: 536335 num_examples: 1601 - name: val_9 num_bytes: 267330 num_examples: 798 - name: test_9 num_bytes: 268335 num_examples: 801 download_size: 3586688 dataset_size: 10720000 - config_name: actg_syn features: - name: setup_key dtype: string - name: scenario dtype: string - name: idx dtype: int64 - name: observed_time dtype: float64 - name: event dtype: float64 - name: T0 dtype: float64 - name: T1 dtype: float64 - name: T dtype: float64 - name: C dtype: float64 - name: W dtype: float64 - name: true_cate dtype: float64 - name: age dtype: int64 - name: wtkg dtype: float64 - name: hemo dtype: int64 - name: homo dtype: int64 - name: drugs dtype: int64 - name: karnof dtype: int64 - name: oprior dtype: int64 - name: z30 dtype: int64 - name: zprior dtype: int64 - name: preanti dtype: int64 - name: race dtype: int64 - name: gender dtype: int64 - name: str2 dtype: int64 - name: strat dtype: int64 - name: symptom dtype: int64 - name: treat dtype: int64 - name: offtrt dtype: int64 - name: cd40 dtype: int64 - name: cd420 dtype: int64 - name: cd496 dtype: float64 - name: r dtype: int64 - name: cd80 dtype: int64 - name: cd820 dtype: int64 - name: p_surv_t25_w0 dtype: float64 - name: p_surv_t50_w0 dtype: float64 - name: p_surv_t75_w0 dtype: float64 - name: p_surv_t25_w1 dtype: float64 - name: p_surv_t50_w1 dtype: float64 - name: p_surv_t75_w1 dtype: float64 - name: summary_json dtype: string splits: - name: train num_bytes: 3599937 num_examples: 2139 download_size: 333223 dataset_size: 3599937 - config_name: actg_syn_repeats features: - name: repeat_key dtype: string - name: idx dtype: int64 - name: random_idx0 dtype: int64 - name: random_idx1 dtype: int64 - name: random_idx2 dtype: int64 - name: random_idx3 dtype: int64 - name: random_idx4 dtype: int64 - name: random_idx5 dtype: int64 - name: random_idx6 dtype: int64 - name: random_idx7 dtype: int64 - name: random_idx8 dtype: int64 - name: random_idx9 dtype: int64 splits: - name: train num_bytes: 211761 num_examples: 2139 download_size: 135284 dataset_size: 211761 - config_name: actg_syn_splits features: - name: setup_key dtype: string - name: scenario dtype: string - name: idx dtype: int64 - name: observed_time dtype: float64 - name: event dtype: float64 - name: T0 dtype: float64 - name: T1 dtype: float64 - name: T dtype: float64 - name: C dtype: float64 - name: W dtype: float64 - name: true_cate dtype: float64 - name: age dtype: int64 - name: wtkg dtype: float64 - name: hemo dtype: int64 - name: homo dtype: int64 - name: drugs dtype: int64 - name: karnof dtype: int64 - name: oprior dtype: int64 - name: z30 dtype: int64 - name: zprior dtype: int64 - name: preanti dtype: int64 - name: race dtype: int64 - name: gender dtype: int64 - name: str2 dtype: int64 - name: strat dtype: int64 - name: symptom dtype: int64 - name: treat dtype: int64 - name: offtrt dtype: int64 - name: cd40 dtype: int64 - name: cd420 dtype: int64 - name: cd496 dtype: float64 - name: r dtype: int64 - name: cd80 dtype: int64 - name: cd820 dtype: int64 - name: p_surv_t25_w0 dtype: float64 - name: p_surv_t50_w0 dtype: float64 - name: p_surv_t75_w0 dtype: float64 - name: p_surv_t25_w1 dtype: float64 - name: p_surv_t50_w1 dtype: float64 - name: p_surv_t75_w1 dtype: float64 splits: - name: train_0 num_bytes: 352770 num_examples: 1069 - name: val_0 num_bytes: 176220 num_examples: 534 - name: test_0 num_bytes: 176880 num_examples: 536 - name: train_1 num_bytes: 352770 num_examples: 1069 - name: val_1 num_bytes: 176220 num_examples: 534 - name: test_1 num_bytes: 176880 num_examples: 536 - name: train_2 num_bytes: 352770 num_examples: 1069 - name: val_2 num_bytes: 176220 num_examples: 534 - name: test_2 num_bytes: 176880 num_examples: 536 - name: train_3 num_bytes: 352770 num_examples: 1069 - name: val_3 num_bytes: 176220 num_examples: 534 - name: test_3 num_bytes: 176880 num_examples: 536 - name: train_4 num_bytes: 352770 num_examples: 1069 - name: val_4 num_bytes: 176220 num_examples: 534 - name: test_4 num_bytes: 176880 num_examples: 536 - name: train_5 num_bytes: 352770 num_examples: 1069 - name: val_5 num_bytes: 176220 num_examples: 534 - name: test_5 num_bytes: 176880 num_examples: 536 - name: train_6 num_bytes: 352770 num_examples: 1069 - name: val_6 num_bytes: 176220 num_examples: 534 - name: test_6 num_bytes: 176880 num_examples: 536 - name: train_7 num_bytes: 352770 num_examples: 1069 - name: val_7 num_bytes: 176220 num_examples: 534 - name: test_7 num_bytes: 176880 num_examples: 536 - name: train_8 num_bytes: 352770 num_examples: 1069 - name: val_8 num_bytes: 176220 num_examples: 534 - name: test_8 num_bytes: 176880 num_examples: 536 - name: train_9 num_bytes: 352770 num_examples: 1069 - name: val_9 num_bytes: 176220 num_examples: 534 - name: test_9 num_bytes: 176880 num_examples: 536 download_size: 3630799 dataset_size: 7058700 - config_name: synthetic features: - name: setup_key dtype: string - name: scenario dtype: string - name: id dtype: int64 - name: observed_time dtype: float64 - name: event dtype: int64 - name: W dtype: int64 - name: X1 dtype: float64 - name: X2 dtype: float64 - name: X3 dtype: float64 - name: X4 dtype: float64 - name: X5 dtype: float64 - name: U1 dtype: float64 - name: U2 dtype: float64 - name: T0 dtype: float64 - name: T1 dtype: float64 - name: T dtype: float64 - name: C dtype: float64 - name: summary_json dtype: string - name: metadata_json dtype: string splits: - name: setups num_bytes: 2276200000 num_examples: 2000000 - name: train num_bytes: 2276200000 num_examples: 2000000 download_size: 1789550600 dataset_size: 4552400000 - config_name: synthetic_repeats features: - name: repeat_key dtype: string - name: idx dtype: int64 - name: random_idx0 dtype: int64 - name: random_idx1 dtype: int64 - name: random_idx2 dtype: int64 - name: random_idx3 dtype: int64 - name: random_idx4 dtype: int64 - name: random_idx5 dtype: int64 - name: random_idx6 dtype: int64 - name: random_idx7 dtype: int64 - name: random_idx8 dtype: int64 - name: random_idx9 dtype: int64 splits: - name: train num_bytes: 4950000 num_examples: 50000 download_size: 3447166 dataset_size: 4950000 - config_name: synthetic_splits features: - name: setup_key dtype: string - name: scenario dtype: string - name: id dtype: int64 - name: observed_time dtype: float64 - name: event dtype: int64 - name: W dtype: int64 - name: X1 dtype: float64 - name: X2 dtype: float64 - name: X3 dtype: float64 - name: X4 dtype: float64 - name: X5 dtype: float64 - name: U1 dtype: float64 - name: U2 dtype: float64 - name: T0 dtype: float64 - name: T1 dtype: float64 - name: T dtype: float64 - name: C dtype: float64 splits: - name: train_0 num_bytes: 29350000 num_examples: 200000 - name: val_0 num_bytes: 14675000 num_examples: 100000 - name: test_0 num_bytes: 14675000 num_examples: 100000 - name: train_1 num_bytes: 29350000 num_examples: 200000 - name: val_1 num_bytes: 14675000 num_examples: 100000 - name: test_1 num_bytes: 14675000 num_examples: 100000 - name: train_2 num_bytes: 29350000 num_examples: 200000 - name: val_2 num_bytes: 14675000 num_examples: 100000 - name: test_2 num_bytes: 14675000 num_examples: 100000 - name: train_3 num_bytes: 29350000 num_examples: 200000 - name: val_3 num_bytes: 14675000 num_examples: 100000 - name: test_3 num_bytes: 14675000 num_examples: 100000 - name: train_4 num_bytes: 29350000 num_examples: 200000 - name: val_4 num_bytes: 14675000 num_examples: 100000 - name: test_4 num_bytes: 14675000 num_examples: 100000 - name: train_5 num_bytes: 29350000 num_examples: 200000 - name: val_5 num_bytes: 14675000 num_examples: 100000 - name: test_5 num_bytes: 14675000 num_examples: 100000 - name: train_6 num_bytes: 29350000 num_examples: 200000 - name: val_6 num_bytes: 14675000 num_examples: 100000 - name: test_6 num_bytes: 14675000 num_examples: 100000 - name: train_7 num_bytes: 29350000 num_examples: 200000 - name: val_7 num_bytes: 14675000 num_examples: 100000 - name: test_7 num_bytes: 14675000 num_examples: 100000 - name: train_8 num_bytes: 29350000 num_examples: 200000 - name: val_8 num_bytes: 14675000 num_examples: 100000 - name: test_8 num_bytes: 14675000 num_examples: 100000 - name: train_9 num_bytes: 29350000 num_examples: 200000 - name: val_9 num_bytes: 14675000 num_examples: 100000 - name: test_9 num_bytes: 14675000 num_examples: 100000 download_size: 104017227 dataset_size: 587000000 - config_name: twin features: - name: setup_key dtype: string - name: scenario dtype: string - name: idx dtype: int64 - name: observed_time dtype: float64 - name: event dtype: int64 - name: T0 dtype: int64 - name: T1 dtype: int64 - name: T dtype: int64 - name: C dtype: float64 - name: W dtype: int64 - name: true_cate dtype: int64 - name: anemia dtype: int64 - name: cardiac dtype: int64 - name: lung dtype: int64 - name: diabetes dtype: int64 - name: herpes dtype: int64 - name: hydra dtype: int64 - name: hemo dtype: int64 - name: chyper dtype: int64 - name: phyper dtype: int64 - name: eclamp dtype: int64 - name: incervix dtype: int64 - name: pre4000 dtype: int64 - name: preterm dtype: int64 - name: renal dtype: int64 - name: rh dtype: int64 - name: uterine dtype: int64 - name: othermr dtype: int64 - name: gestat dtype: int64 - name: dmage dtype: int64 - name: dmeduc dtype: int64 - name: dmar dtype: int64 - name: nprevist dtype: int64 - name: adequacy dtype: int64 - name: dtotord dtype: int64 - name: cigar dtype: int64 - name: drink dtype: int64 - name: wtgain dtype: int64 - name: pldel_2 dtype: int64 - name: pldel_3 dtype: int64 - name: pldel_4 dtype: int64 - name: pldel_5 dtype: int64 - name: resstatb_2 dtype: int64 - name: resstatb_3 dtype: int64 - name: resstatb_4 dtype: int64 - name: mpcb_1 dtype: int64 - name: mpcb_2 dtype: int64 - name: mpcb_3 dtype: int64 - name: mpcb_4 dtype: int64 - name: mpcb_5 dtype: int64 - name: mpcb_6 dtype: int64 - name: mpcb_7 dtype: int64 - name: mpcb_8 dtype: int64 - name: mpcb_9 dtype: int64 - name: summary_json dtype: string splits: - name: train num_bytes: 23187600 num_examples: 22800 download_size: 474020 dataset_size: 23187600 - config_name: twin_repeats features: - name: repeat_key dtype: string - name: idx dtype: int64 - name: random_idx0 dtype: int64 - name: random_idx1 dtype: int64 - name: random_idx2 dtype: int64 - name: random_idx3 dtype: int64 - name: random_idx4 dtype: int64 - name: random_idx5 dtype: int64 - name: random_idx6 dtype: int64 - name: random_idx7 dtype: int64 - name: random_idx8 dtype: int64 - name: random_idx9 dtype: int64 splits: - name: train num_bytes: 1128600 num_examples: 11400 download_size: 728359 dataset_size: 1128600 - config_name: twin_splits features: - name: setup_key dtype: string - name: scenario dtype: string - name: idx dtype: int64 - name: observed_time dtype: float64 - name: event dtype: int64 - name: T0 dtype: int64 - name: T1 dtype: int64 - name: T dtype: int64 - name: C dtype: float64 - name: W dtype: int64 - name: true_cate dtype: int64 - name: anemia dtype: int64 - name: cardiac dtype: int64 - name: lung dtype: int64 - name: diabetes dtype: int64 - name: herpes dtype: int64 - name: hydra dtype: int64 - name: hemo dtype: int64 - name: chyper dtype: int64 - name: phyper dtype: int64 - name: eclamp dtype: int64 - name: incervix dtype: int64 - name: pre4000 dtype: int64 - name: preterm dtype: int64 - name: renal dtype: int64 - name: rh dtype: int64 - name: uterine dtype: int64 - name: othermr dtype: int64 - name: gestat dtype: int64 - name: dmage dtype: int64 - name: dmeduc dtype: int64 - name: dmar dtype: int64 - name: nprevist dtype: int64 - name: adequacy dtype: int64 - name: dtotord dtype: int64 - name: cigar dtype: int64 - name: drink dtype: int64 - name: wtgain dtype: int64 - name: pldel_2 dtype: int64 - name: pldel_3 dtype: int64 - name: pldel_4 dtype: int64 - name: pldel_5 dtype: int64 - name: resstatb_2 dtype: int64 - name: resstatb_3 dtype: int64 - name: resstatb_4 dtype: int64 - name: mpcb_1 dtype: int64 - name: mpcb_2 dtype: int64 - name: mpcb_3 dtype: int64 - name: mpcb_4 dtype: int64 - name: mpcb_5 dtype: int64 - name: mpcb_6 dtype: int64 - name: mpcb_7 dtype: int64 - name: mpcb_8 dtype: int64 - name: mpcb_9 dtype: int64 splits: - name: train_0 num_bytes: 5021700 num_examples: 11400 - name: val_0 num_bytes: 2510850 num_examples: 5700 - name: test_0 num_bytes: 2510850 num_examples: 5700 - name: train_1 num_bytes: 5021700 num_examples: 11400 - name: val_1 num_bytes: 2510850 num_examples: 5700 - name: test_1 num_bytes: 2510850 num_examples: 5700 - name: train_2 num_bytes: 5021700 num_examples: 11400 - name: val_2 num_bytes: 2510850 num_examples: 5700 - name: test_2 num_bytes: 2510850 num_examples: 5700 - name: train_3 num_bytes: 5021700 num_examples: 11400 - name: val_3 num_bytes: 2510850 num_examples: 5700 - name: test_3 num_bytes: 2510850 num_examples: 5700 - name: train_4 num_bytes: 5021700 num_examples: 11400 - name: val_4 num_bytes: 2510850 num_examples: 5700 - name: test_4 num_bytes: 2510850 num_examples: 5700 - name: train_5 num_bytes: 5021700 num_examples: 11400 - name: val_5 num_bytes: 2510850 num_examples: 5700 - name: test_5 num_bytes: 2510850 num_examples: 5700 - name: train_6 num_bytes: 5021700 num_examples: 11400 - name: val_6 num_bytes: 2510850 num_examples: 5700 - name: test_6 num_bytes: 2510850 num_examples: 5700 - name: train_7 num_bytes: 5021700 num_examples: 11400 - name: val_7 num_bytes: 2510850 num_examples: 5700 - name: test_7 num_bytes: 2510850 num_examples: 5700 - name: train_8 num_bytes: 5021700 num_examples: 11400 - name: val_8 num_bytes: 2510850 num_examples: 5700 - name: test_8 num_bytes: 2510850 num_examples: 5700 - name: train_9 num_bytes: 5021700 num_examples: 11400 - name: val_9 num_bytes: 2510850 num_examples: 5700 - name: test_9 num_bytes: 2510850 num_examples: 5700 download_size: 4822219 dataset_size: 100434000 configs: - config_name: actgHC data_files: - split: train path: actgHC/train-* - config_name: actgHC_repeats data_files: - split: train path: actgHC_repeats/train-* - config_name: actgHC_splits data_files: - split: train_0 path: actgHC_splits/train_0-* - split: val_0 path: actgHC_splits/val_0-* - split: test_0 path: actgHC_splits/test_0-* - split: train_1 path: actgHC_splits/train_1-* - split: val_1 path: actgHC_splits/val_1-* - split: test_1 path: actgHC_splits/test_1-* - split: train_2 path: actgHC_splits/train_2-* - split: val_2 path: actgHC_splits/val_2-* - split: test_2 path: actgHC_splits/test_2-* - split: train_3 path: actgHC_splits/train_3-* - split: val_3 path: actgHC_splits/val_3-* - split: test_3 path: actgHC_splits/test_3-* - split: train_4 path: actgHC_splits/train_4-* - split: val_4 path: actgHC_splits/val_4-* - split: test_4 path: actgHC_splits/test_4-* - split: train_5 path: actgHC_splits/train_5-* - split: val_5 path: actgHC_splits/val_5-* - split: test_5 path: actgHC_splits/test_5-* - split: train_6 path: actgHC_splits/train_6-* - split: val_6 path: actgHC_splits/val_6-* - split: test_6 path: actgHC_splits/test_6-* - split: train_7 path: actgHC_splits/train_7-* - split: val_7 path: actgHC_splits/val_7-* - split: test_7 path: actgHC_splits/test_7-* - split: train_8 path: actgHC_splits/train_8-* - split: val_8 path: actgHC_splits/val_8-* - split: test_8 path: actgHC_splits/test_8-* - split: train_9 path: actgHC_splits/train_9-* - split: val_9 path: actgHC_splits/val_9-* - split: test_9 path: actgHC_splits/test_9-* - config_name: actgLC data_files: - split: train path: actgLC/train-* - config_name: actgLC_repeats data_files: - split: train path: actgLC_repeats/train-* - config_name: actgLC_splits data_files: - split: train_0 path: actgLC_splits/train_0-* - split: val_0 path: actgLC_splits/val_0-* - split: test_0 path: actgLC_splits/test_0-* - split: train_1 path: actgLC_splits/train_1-* - split: val_1 path: actgLC_splits/val_1-* - split: test_1 path: actgLC_splits/test_1-* - split: train_2 path: actgLC_splits/train_2-* - split: val_2 path: actgLC_splits/val_2-* - split: test_2 path: actgLC_splits/test_2-* - split: train_3 path: actgLC_splits/train_3-* - split: val_3 path: actgLC_splits/val_3-* - split: test_3 path: actgLC_splits/test_3-* - split: train_4 path: actgLC_splits/train_4-* - split: val_4 path: actgLC_splits/val_4-* - split: test_4 path: actgLC_splits/test_4-* - split: train_5 path: actgLC_splits/train_5-* - split: val_5 path: actgLC_splits/val_5-* - split: test_5 path: actgLC_splits/test_5-* - split: train_6 path: actgLC_splits/train_6-* - split: val_6 path: actgLC_splits/val_6-* - split: test_6 path: actgLC_splits/test_6-* - split: train_7 path: actgLC_splits/train_7-* - split: val_7 path: actgLC_splits/val_7-* - split: test_7 path: actgLC_splits/test_7-* - split: train_8 path: actgLC_splits/train_8-* - split: val_8 path: actgLC_splits/val_8-* - split: test_8 path: actgLC_splits/test_8-* - split: train_9 path: actgLC_splits/train_9-* - split: val_9 path: actgLC_splits/val_9-* - split: test_9 path: actgLC_splits/test_9-* - config_name: actg_syn data_files: - split: train path: actg_syn/train-* - config_name: actg_syn_repeats data_files: - split: train path: actg_syn_repeats/train-* - config_name: actg_syn_splits data_files: - split: train_0 path: actg_syn_splits/train_0-* - split: val_0 path: actg_syn_splits/val_0-* - split: test_0 path: actg_syn_splits/test_0-* - split: train_1 path: actg_syn_splits/train_1-* - split: val_1 path: actg_syn_splits/val_1-* - split: test_1 path: actg_syn_splits/test_1-* - split: train_2 path: actg_syn_splits/train_2-* - split: val_2 path: actg_syn_splits/val_2-* - split: test_2 path: actg_syn_splits/test_2-* - split: train_3 path: actg_syn_splits/train_3-* - split: val_3 path: actg_syn_splits/val_3-* - split: test_3 path: actg_syn_splits/test_3-* - split: train_4 path: actg_syn_splits/train_4-* - split: val_4 path: actg_syn_splits/val_4-* - split: test_4 path: actg_syn_splits/test_4-* - split: train_5 path: actg_syn_splits/train_5-* - split: val_5 path: actg_syn_splits/val_5-* - split: test_5 path: actg_syn_splits/test_5-* - split: train_6 path: actg_syn_splits/train_6-* - split: val_6 path: actg_syn_splits/val_6-* - split: test_6 path: actg_syn_splits/test_6-* - split: train_7 path: actg_syn_splits/train_7-* - split: val_7 path: actg_syn_splits/val_7-* - split: test_7 path: actg_syn_splits/test_7-* - split: train_8 path: actg_syn_splits/train_8-* - split: val_8 path: actg_syn_splits/val_8-* - split: test_8 path: actg_syn_splits/test_8-* - split: train_9 path: actg_syn_splits/train_9-* - split: val_9 path: actg_syn_splits/val_9-* - split: test_9 path: actg_syn_splits/test_9-* - config_name: synthetic data_files: - split: setups path: synthetic/setups-* - split: train path: synthetic/train-* - config_name: synthetic_repeats data_files: - split: train path: synthetic_repeats/train-* - config_name: synthetic_splits data_files: - split: train_0 path: synthetic_splits/train_0-* - split: val_0 path: synthetic_splits/val_0-* - split: test_0 path: synthetic_splits/test_0-* - split: train_1 path: synthetic_splits/train_1-* - split: val_1 path: synthetic_splits/val_1-* - split: test_1 path: synthetic_splits/test_1-* - split: train_2 path: synthetic_splits/train_2-* - split: val_2 path: synthetic_splits/val_2-* - split: test_2 path: synthetic_splits/test_2-* - split: train_3 path: synthetic_splits/train_3-* - split: val_3 path: synthetic_splits/val_3-* - split: test_3 path: synthetic_splits/test_3-* - split: train_4 path: synthetic_splits/train_4-* - split: val_4 path: synthetic_splits/val_4-* - split: test_4 path: synthetic_splits/test_4-* - split: train_5 path: synthetic_splits/train_5-* - split: val_5 path: synthetic_splits/val_5-* - split: test_5 path: synthetic_splits/test_5-* - split: train_6 path: synthetic_splits/train_6-* - split: val_6 path: synthetic_splits/val_6-* - split: test_6 path: synthetic_splits/test_6-* - split: train_7 path: synthetic_splits/train_7-* - split: val_7 path: synthetic_splits/val_7-* - split: test_7 path: synthetic_splits/test_7-* - split: train_8 path: synthetic_splits/train_8-* - split: val_8 path: synthetic_splits/val_8-* - split: test_8 path: synthetic_splits/test_8-* - split: train_9 path: synthetic_splits/train_9-* - split: val_9 path: synthetic_splits/val_9-* - split: test_9 path: synthetic_splits/test_9-* - config_name: twin data_files: - split: train path: twin/train-* - config_name: twin_repeats data_files: - split: train path: twin_repeats/train-* - config_name: twin_splits data_files: - split: train_0 path: twin_splits/train_0-* - split: val_0 path: twin_splits/val_0-* - split: test_0 path: twin_splits/test_0-* - split: train_1 path: twin_splits/train_1-* - split: val_1 path: twin_splits/val_1-* - split: test_1 path: twin_splits/test_1-* - split: train_2 path: twin_splits/train_2-* - split: val_2 path: twin_splits/val_2-* - split: test_2 path: twin_splits/test_2-* - split: train_3 path: twin_splits/train_3-* - split: val_3 path: twin_splits/val_3-* - split: test_3 path: twin_splits/test_3-* - split: train_4 path: twin_splits/train_4-* - split: val_4 path: twin_splits/val_4-* - split: test_4 path: twin_splits/test_4-* - split: train_5 path: twin_splits/train_5-* - split: val_5 path: twin_splits/val_5-* - split: test_5 path: twin_splits/test_5-* - split: train_6 path: twin_splits/train_6-* - split: val_6 path: twin_splits/val_6-* - split: test_6 path: twin_splits/test_6-* - split: train_7 path: twin_splits/train_7-* - split: val_7 path: twin_splits/val_7-* - split: test_7 path: twin_splits/test_7-* - split: train_8 path: twin_splits/train_8-* - split: val_8 path: twin_splits/val_8-* - split: test_8 path: twin_splits/test_8-* - split: train_9 path: twin_splits/train_9-* - split: val_9 path: twin_splits/val_9-* - split: test_9 path: twin_splits/test_9-* --- # SurvHTE-Bench: A Benchmark for Heterogeneous Treatment Effect Estimation in Survival Analysis **Paper:** [ICLR 2026 — SurvHTE-Bench: A Benchmark for Heterogeneous Treatment Effect Estimation in Survival Analysis](https://huggingface.co/papers/2603.05483) **GitHub:** https://github.com/Shahriarnz14/SurvHTE-Bench --- ## Overview SurvHTE-Bench is a benchmark for **heterogeneous treatment effect (HTE) estimation under right-censored survival outcomes**. The benchmark addresses an important gap at the intersection of causal inference and survival analysis. While heterogeneous treatment effect estimation has been widely studied in fully observed outcome settings, systematic evaluation in **time-to-event data with censoring** has been largely missing. SurvHTE-Bench provides a unified framework for evaluating survival HTE estimators across: - **Synthetic datasets** with known ground-truth treatment effects - **Semi-synthetic datasets** combining real covariates with simulated treatments and outcomes - **Real-world datasets** including a twin birth dataset (with ground-truth counterfactual outcomes) and an HIV clinical trial dataset Across these datasets, the benchmark evaluates **53 estimator variants** spanning three major methodological families: - Outcome imputation approaches - Direct survival causal methods - Survival meta-learners The benchmark focuses primarily on **Conditional Average Treatment Effects (CATE)** defined using **Restricted Mean Survival Time (RMST)** as the survival estimand. --- ## Datasets The benchmark includes five dataset groups spanning the full data-generation spectrum. ### 1. `synthetic` — Fully Synthetic The synthetic benchmark consists of **40 datasets**, constructed by crossing: - **8 causal configurations** (different treatment assignment mechanisms, confounding structures, positivity violations, and censoring mechanisms) - **5 survival scenarios** (different survival distributions and censoring regimes) Each dataset contains up to **50,000 samples** with: - **5 covariates** independently sampled from Uniform(0,1) - binary treatment `W` - observed time `observed_time` - event indicator `event` - potential survival times `T0` and `T1` Because both potential outcomes are generated, **ground-truth individual treatment effects are available**. The causal configurations include randomized controlled trials and observational settings with violations such as unmeasured confounding, lack of positivity, and informative censoring. ### 2. `actg_syn` — Semi-Synthetic ACTG Dataset Semi-synthetic datasets constructed from the **ACTG 175 HIV clinical trial**, which contains **2,139 patients**. - Covariates are real patient features from the trial. - Treatment assignments and survival outcomes are **simulated** to generate known treatment effects. This preserves realistic covariate distributions while enabling controlled evaluation. ### 3. `twin` — Twin Birth Dataset A real-world dataset derived from the **Twin Births dataset**, containing **11,400 twin pairs**. The twin structure allows near-counterfactual evaluation: for each pair, one twin is treated and the other is untreated. Treatment corresponds to **being the heavier twin**, and the outcome is **time to mortality**. ### 4. `actgHC` — ACTG High-Censoring Variant A version of the ACTG dataset with **high censoring rates**, containing approximately **1,054–1,093 samples** depending on the trial arm. The dataset includes multiple time/event pairs (`t0/e0` … `t9/e9`) representing repeated survival observations. ### 5. `actgLC` — ACTG Low-Censoring Variant A lower-censoring version of the ACTG dataset. The structure mirrors `actgHC`, but censoring rates are substantially lower. ### 6. `mimic_syn` — Semi-Synthetic MIMIC-IV Datasets The benchmark also includes **semi-synthetic datasets derived from covariates in the MIMIC-IV ICU database**. In the paper, we construct **nine MIMIC-based semi-synthetic datasets (MIMIC-i – MIMIC-ix)** using real patient covariates from MIMIC-IV while simulating treatment assignments and survival outcomes. These datasets are designed to capture realistic covariate structure while enabling controlled evaluation with known ground-truth treatment effects. The datasets cover multiple regimes: - **MIMIC-i – MIMIC-v:** varying censoring severity (approximately 53%–88%) under covariate-independent treatment assignment. - **MIMIC-vi – MIMIC-ix:** covariate-dependent treatment assignment with more complex nonlinear outcome and censoring mechanisms. Due to the **MIMIC-IV data usage agreement**, we cannot redistribute the original data or any datasets derived directly from it through this repository or the HuggingFace dataset. Researchers must obtain access to MIMIC-IV through PhysioNet: https://physionet.org/content/mimiciv/ After obtaining access, the semi-synthetic datasets used in our experiments can be reproduced using the notebook provided in the repository: https://github.com/Shahriarnz14/SurvHTE-Bench/blob/main/data/semi-synthetic/generate_mimic_semi_synthetic.ipynb --- ## HuggingFace Configuration Layout Each dataset group is split into three HuggingFace configurations: | Config name | Split(s) | Contents | |---|---|---| | `{name}` | `train` | Full data with metadata | | `{name}_repeats` | `train` | Random index permutations used for repeated splits | | `{name}_splits` | `train_0`…`train_9`, `val_0`…`val_9`, `test_0`…`test_9` | Pre-computed splits for repeated experiments | So the full list of configs is: `synthetic`, `synthetic_repeats`, `synthetic_splits`, `actg_syn`, `actg_syn_repeats`, `actg_syn_splits`, `twin`, `twin_repeats`, `twin_splits`, `actgHC`, `actgHC_repeats`, `actgHC_splits`, `actgLC`, `actgLC_repeats`, `actgLC_splits` --- ## Loading the Data We provide a ready-to-use loader at [`data_utils/hf_load.py`](https://github.com/Shahriarnz14/SurvHTE-Bench/blob/main/data_utils/hf_load.py) in the GitHub repository. Install dependencies first: ```bash pip install datasets pandas numpy ``` ### Interface 1 — `load_data`: Full Dataset (mirrors local API) Reconstructs `experiment_setups` and `experiment_repeat_setups` identically to the original local data loader. ```python from data_utils.hf_load import load_data experiment_setups, experiment_repeat_setups = load_data(dataset_name="synthetic") ``` `experiment_setups` is a nested dict: ``` experiment_setups[setup_key][scenario] = { "dataset": pd.DataFrame, # all covariates + outcome columns "summary": dict, # summary statistics "metadata": dict, # (synthetic only) DGP metadata } ``` `experiment_repeat_setups` contains the pre-computed random index permutations used to generate reproducible train/val/test splits. For `actgHC`/`actgLC` it is a `{setup_key: DataFrame}` dict; for all other datasets it is a single shared `DataFrame`. Supported `dataset_name` values: `"synthetic"`, `"actg_syn"`, `"twin"`, `"actgHC"`, `"actgLC"`. ### Interface 2 — `load_splits`: Pre-Split Arrays (drop-in for experiment loop) Returns arrays already split into train/val/test for each configuration, scenario, and repeat index — ready to pass directly into model training. ```python from data_utils.hf_load import load_splits split_dict = load_splits(dataset_name="synthetic") ``` The returned structure is: ``` split_dict[config_name][scenario_key][rand_idx]["train" | "val" | "test"] = (X, W, Y, cate_true) ``` where: - `X` — covariate matrix `(n, d)` as `np.ndarray` - `W` — treatment vector `(n,)` as `np.ndarray` - `Y` — outcome matrix `(n, 2)` containing `[observed_time, event]` (or all `t/e` columns for `actgHC`) - `cate_true` — ground-truth CATE `(n,)` (or proxy) **Example — accessing a specific split:** ```python config_name = "RCT-50" # setup key scenario_key = "Scenario_A" # scenario rand_idx = 0 # repeat index (0–9) X_train, W_train, Y_train, cate_true_train = split_dict[config_name][scenario_key][rand_idx]["train"] X_val, W_val, Y_val, cate_true_val = split_dict[config_name][scenario_key][rand_idx]["val"] X_test, W_test, Y_test, cate_true_test = split_dict[config_name][scenario_key][rand_idx]["test"] ``` **Example — iterating the full experiment loop:** ```python results = load_splits(dataset_name="synthetic") for config_name, scenarios in results.items(): for scenario_key, repeats in scenarios.items(): for rand_idx in range(10): X_tr, W_tr, Y_tr, cate_tr = repeats[rand_idx]["train"] X_te, W_te, Y_te, cate_te = repeats[rand_idx]["test"] # ... fit model, evaluate ... ``` --- ## Evaluation The benchmark evaluates heterogeneous treatment effect estimators using metrics derived from the **true Conditional Average Treatment Effect (CATE)**. Primary evaluation metrics include: - **CATE Root Mean Square Error (RMSE)** Measures the error between estimated and true individual treatment effects. - **ATE Bias** Measures the deviation of the estimated average treatment effect from the true population ATE. Additional auxiliary metrics are used to analyze component performance: - **Imputation accuracy** (for methods using survival time imputation) - **Regression or survival model performance**, such as MAE or time-dependent C-index All experiments are averaged over **10 repeated train/validation/test splits**. --- ## Repository Full code and experiment scripts are available at: https://github.com/Shahriarnz14/SurvHTE-Bench --- ## Citation If you use SurvHTE-Bench in your research, please cite: ```bibtex @inproceedings{noroozizadeh2026survhte, title={SurvHTE-Bench: A Benchmark for Heterogeneous Treatment Effect Estimation in Survival Analysis}, author={Noroozizadeh, Shahriar and Shen, Xiaobin and Weiss, Jeremy and Chen, George H.}, booktitle={International Conference on Learning Representations (ICLR)}, year={2026} } ``` --- ## License This dataset is released under the **Creative Commons Attribution 4.0 International (CC BY 4.0)** license.