survival 1.1.19__cp314-cp314-manylinux_2_39_x86_64.whl → 1.1.32__cp314-cp314-manylinux_2_39_x86_64.whl

survival/__init__.py

@@ -1,5 +1,15 @@
-from .survival import *
-
-__doc__ = survival.__doc__
-if hasattr(survival, "__all__"):
-    __all__ = survival.__all__
+from survival._survival import *  # noqa: F401, F403
+from survival.sklearn_compat import (  # noqa: F401
+    AFTEstimator,
+    CoxPHEstimator,
+    GradientBoostSurvivalEstimator,
+    StreamingAFTEstimator,
+    StreamingCoxPHEstimator,
+    StreamingGradientBoostSurvivalEstimator,
+    StreamingMixin,
+    StreamingSurvivalForestEstimator,
+    SurvivalForestEstimator,
+    iter_chunks,
+    predict_large_dataset,
+    survival_curves_to_disk,
+)

survival/_survival.pyi

@@ -0,0 +1,660 @@
+from typing import Any
+
+class GBSurvLoss:
+    AFT: GBSurvLoss
+    CoxPH: GBSurvLoss
+    Huber: GBSurvLoss
+    SquaredError: GBSurvLoss
+
+class SplitRule:
+    Conservation: SplitRule
+    Hellinger: SplitRule
+    LogRank: SplitRule
+    LogRankScore: SplitRule
+
+class DistributionType:
+    ExtremeValue: DistributionType
+    Weibull: DistributionType
+    Gaussian: DistributionType
+    Logistic: DistributionType
+    LogNormal: DistributionType
+    LogLogistic: DistributionType
+
+class DimType:
+    Age: DimType
+    Continuous: DimType
+    Factor: DimType
+    Year: DimType
+
+class ReliabilityScale:
+    ClogLog: ReliabilityScale
+    Cumhaz: ReliabilityScale
+    LogLogistic: ReliabilityScale
+    Probit: ReliabilityScale
+    Surv: ReliabilityScale
+
+class GradientBoostSurvivalConfig:
+    n_estimators: int
+    learning_rate: float
+    max_depth: int
+    min_samples_split: int
+    min_samples_leaf: int
+    subsample: float
+    max_features: int | None
+    seed: int | None
+    loss: GBSurvLoss
+    dropout_rate: float
+
+    def __init__(
+        self,
+        n_estimators: int = 100,
+        learning_rate: float = 0.1,
+        max_depth: int = 3,
+        min_samples_split: int = 10,
+        min_samples_leaf: int = 5,
+        subsample: float = 1.0,
+        max_features: int | None = None,
+        seed: int | None = None,
+        loss: GBSurvLoss | None = None,
+        dropout_rate: float = 0.0,
+    ) -> None: ...
+
+class GradientBoostSurvival:
+    @property
+    def n_estimators(self) -> int: ...
+    @property
+    def learning_rate(self) -> float: ...
+    @property
+    def unique_times(self) -> list[float]: ...
+    @property
+    def feature_importance(self) -> list[float]: ...
+    @property
+    def baseline_hazard(self) -> list[float]: ...
+    @property
+    def train_loss(self) -> list[float]: ...
+    @staticmethod
+    def fit(
+        x: list[float],
+        n: int,
+        p: int,
+        time: list[float],
+        status: list[int],
+        config: GradientBoostSurvivalConfig,
+    ) -> GradientBoostSurvival: ...
+    def predict_risk(self, x_new: list[float], n_new: int) -> list[float]: ...
+    def predict_survival(self, x_new: list[float], n_new: int) -> list[list[float]]: ...
+    def predict_cumulative_hazard(self, x_new: list[float], n_new: int) -> list[list[float]]: ...
+    def predict_survival_time(
+        self, x_new: list[float], n_new: int, percentile: float = 0.5
+    ) -> list[float | None]: ...
+    def predict_median_survival_time(
+        self, x_new: list[float], n_new: int
+    ) -> list[float | None]: ...
+
+class SurvivalForestConfig:
+    n_trees: int
+    max_depth: int | None
+    min_node_size: int
+    mtry: int | None
+    sample_fraction: float
+    seed: int | None
+    oob_error: bool
+    split_rule: SplitRule
+    n_random_splits: int
+
+    def __init__(
+        self,
+        n_trees: int = 500,
+        max_depth: int | None = None,
+        min_node_size: int = 15,
+        mtry: int | None = None,
+        sample_fraction: float = 0.632,
+        seed: int | None = None,
+        oob_error: bool = True,
+        split_rule: SplitRule | None = None,
+        n_random_splits: int = 10,
+    ) -> None: ...
+
+class SurvivalForest:
+    @property
+    def n_trees(self) -> int: ...
+    @property
+    def unique_times(self) -> list[float]: ...
+    @property
+    def variable_importance(self) -> list[float]: ...
+    @property
+    def oob_error(self) -> float | None: ...
+    @staticmethod
+    def fit(
+        x: list[float],
+        n: int,
+        p: int,
+        time: list[float],
+        status: list[int],
+        config: SurvivalForestConfig,
+    ) -> SurvivalForest: ...
+    def predict_risk(self, x_new: list[float], n_new: int) -> list[float]: ...
+    def predict_survival(self, x_new: list[float], n_new: int) -> list[list[float]]: ...
+    def predict_cumulative_hazard(self, x_new: list[float], n_new: int) -> list[list[float]]: ...
+    def predict_survival_time(
+        self, x_new: list[float], n_new: int, percentile: float = 0.5
+    ) -> list[float | None]: ...
+    def predict_median_survival_time(
+        self, x_new: list[float], n_new: int
+    ) -> list[float | None]: ...
+
+class Subject:
+    def __init__(
+        self,
+        covariates: list[float],
+        time: float,
+        status: int,
+        weight: float = 1.0,
+    ) -> None: ...
+
+class CoxPHModel:
+    @property
+    def coefficients(self) -> list[float]: ...
+    @property
+    def std_errors(self) -> list[float]: ...
+    @property
+    def vcov(self) -> list[list[float]]: ...
+    @property
+    def log_likelihood(self) -> float: ...
+    @property
+    def n_observations(self) -> int: ...
+    @property
+    def n_events(self) -> int: ...
+    @property
+    def event_times(self) -> list[float]: ...
+    @property
+    def baseline_hazard(self) -> list[float]: ...
+    @property
+    def censoring(self) -> list[int]: ...
+    @staticmethod
+    def new_with_data(
+        covariates: list[list[float]],
+        time: list[float],
+        status: list[int],
+    ) -> CoxPHModel: ...
+    def add_subject(self, subject: Subject) -> None: ...
+    def fit(self, n_iters: int = 20) -> None: ...
+    def predict(self, covariates: list[list[float]]) -> list[float]: ...
+    def predict_survival(self, covariates: list[list[float]]) -> list[list[float]]: ...
+    def predicted_survival_time(
+        self, covariates: list[list[float]], percentile: float = 0.5
+    ) -> list[float | None]: ...
+    def survival_curve(
+        self, covariates: list[list[float]], times: list[float] | None = None
+    ) -> tuple[list[float], list[list[float]]]: ...
+    def cumulative_hazard(self, covariates: list[list[float]]) -> list[list[float]]: ...
+    def risk_scores(self, covariates: list[list[float]]) -> list[float]: ...
+    def hazard_ratios(self) -> list[float]: ...
+    def hazard_ratios_with_ci(
+        self, confidence_level: float = 0.95
+    ) -> list[tuple[float, float, float]]: ...
+    def martingale_residuals(self) -> list[float]: ...
+    def deviance_residuals(self) -> list[float]: ...
+    def dfbeta(self) -> list[list[float]]: ...
+    def brier_score(self, time: float) -> float: ...
+    def restricted_mean_survival_time(self, tau: float) -> float: ...
+    def summary(self) -> str: ...
+    def aic(self) -> float: ...
+    def bic(self) -> float: ...
+    def calculate_baseline_hazard(self) -> None: ...
+    def compute_standard_errors(self) -> None: ...
+    def compute_fisher_information(self) -> list[list[float]]: ...
+
+class SurvFitKMOutput:
+    @property
+    def time(self) -> list[float]: ...
+    @property
+    def surv(self) -> list[float]: ...
+    @property
+    def n_risk(self) -> list[int]: ...
+    @property
+    def n_event(self) -> list[int]: ...
+    @property
+    def n_censor(self) -> list[int]: ...
+    @property
+    def std_err(self) -> list[float]: ...
+    @property
+    def lower(self) -> list[float]: ...
+    @property
+    def upper(self) -> list[float]: ...
+    @property
+    def cumhaz(self) -> list[float]: ...
+
+class KaplanMeierConfig:
+    def __init__(
+        self,
+        conf_level: float = 0.95,
+        conf_type: str = "log-log",
+        se_type: str = "greenwood",
+    ) -> None: ...
+
+class SurvfitKMOptions:
+    def __init__(
+        self,
+        conf_level: float = 0.95,
+        conf_type: str = "log-log",
+        se_type: str = "greenwood",
+        start_time: float | None = None,
+    ) -> None: ...
+
+class NelsonAalenResult:
+    @property
+    def time(self) -> list[float]: ...
+    @property
+    def cumhaz(self) -> list[float]: ...
+    @property
+    def variance(self) -> list[float]: ...
+    @property
+    def lower(self) -> list[float]: ...
+    @property
+    def upper(self) -> list[float]: ...
+    @property
+    def n_risk(self) -> list[int]: ...
+    @property
+    def n_event(self) -> list[int]: ...
+
+class LogRankResult:
+    @property
+    def statistic(self) -> float: ...
+    @property
+    def p_value(self) -> float: ...
+    @property
+    def df(self) -> int: ...
+    @property
+    def observed(self) -> list[float]: ...
+    @property
+    def expected(self) -> list[float]: ...
+    @property
+    def var_observed(self) -> list[float]: ...
+
+class RMSTResult:
+    @property
+    def rmst(self) -> float: ...
+    @property
+    def se(self) -> float: ...
+    @property
+    def lower(self) -> float: ...
+    @property
+    def upper(self) -> float: ...
+    @property
+    def tau(self) -> float: ...
+
+class CalibrationResult:
+    @property
+    def observed(self) -> list[float]: ...
+    @property
+    def predicted(self) -> list[float]: ...
+    @property
+    def n_groups(self) -> int: ...
+    @property
+    def hosmer_lemeshow_stat(self) -> float: ...
+    @property
+    def hosmer_lemeshow_p(self) -> float: ...
+
+class UnoCIndexResult:
+    @property
+    def c_index(self) -> float: ...
+    @property
+    def se(self) -> float: ...
+    @property
+    def lower(self) -> float: ...
+    @property
+    def upper(self) -> float: ...
+    @property
+    def n_pairs(self) -> int: ...
+    @property
+    def n_concordant(self) -> int: ...
+
+class SurvivalFit:
+    @property
+    def coefficients(self) -> list[float]: ...
+    @property
+    def std_errors(self) -> list[float]: ...
+    @property
+    def vcov(self) -> list[list[float]]: ...
+    @property
+    def log_likelihood(self) -> float: ...
+    @property
+    def scale(self) -> float: ...
+    @property
+    def distribution(self) -> str: ...
+    @property
+    def n_observations(self) -> int: ...
+    @property
+    def n_events(self) -> int: ...
+    @property
+    def aic(self) -> float: ...
+    @property
+    def bic(self) -> float: ...
+
+class SurvregConfig:
+    def __init__(
+        self,
+        distribution: DistributionType | None = None,
+        max_iter: int = 100,
+        tol: float = 1e-9,
+    ) -> None: ...
+
+class BootstrapResult:
+    @property
+    def estimate(self) -> float: ...
+    @property
+    def se(self) -> float: ...
+    @property
+    def lower(self) -> float: ...
+    @property
+    def upper(self) -> float: ...
+    @property
+    def bootstrap_estimates(self) -> list[float]: ...
+
+class CVResult:
+    @property
+    def mean_score(self) -> float: ...
+    @property
+    def std_score(self) -> float: ...
+    @property
+    def scores(self) -> list[float]: ...
+    @property
+    def n_folds(self) -> int: ...
+
+class TestResult:
+    @property
+    def statistic(self) -> float: ...
+    @property
+    def p_value(self) -> float: ...
+    @property
+    def df(self) -> int: ...
+
+class BayesianCoxResult:
+    @property
+    def coefficients(self) -> list[float]: ...
+    @property
+    def std_errors(self) -> list[float]: ...
+    @property
+    def credible_intervals(self) -> list[tuple[float, float]]: ...
+    @property
+    def effective_sample_size(self) -> list[float]: ...
+    @property
+    def rhat(self) -> list[float]: ...
+
+class ElasticNetCoxResult:
+    @property
+    def coefficients(self) -> list[float]: ...
+    @property
+    def intercept(self) -> float: ...
+    @property
+    def alpha(self) -> float: ...
+    @property
+    def l1_ratio(self) -> float: ...
+    @property
+    def n_iter(self) -> int: ...
+    @property
+    def nonzero_features(self) -> list[int]: ...
+
+class MixtureCureResult:
+    @property
+    def cure_coefficients(self) -> list[float]: ...
+    @property
+    def survival_coefficients(self) -> list[float]: ...
+    @property
+    def cure_rate(self) -> float: ...
+    @property
+    def log_likelihood(self) -> float: ...
+
+class JointModelResult:
+    @property
+    def longitudinal_coefficients(self) -> list[float]: ...
+    @property
+    def survival_coefficients(self) -> list[float]: ...
+    @property
+    def association(self) -> float: ...
+    @property
+    def log_likelihood(self) -> float: ...
+
+class IPCWResult:
+    @property
+    def weights(self) -> list[float]: ...
+    @property
+    def estimate(self) -> float: ...
+    @property
+    def se(self) -> float: ...
+    @property
+    def lower(self) -> float: ...
+    @property
+    def upper(self) -> float: ...
+
+class GComputationResult:
+    @property
+    def ate(self) -> float: ...
+    @property
+    def se(self) -> float: ...
+    @property
+    def lower(self) -> float: ...
+    @property
+    def upper(self) -> float: ...
+    @property
+    def survival_treated(self) -> list[float]: ...
+    @property
+    def survival_control(self) -> list[float]: ...
+
+class QALYResult:
+    @property
+    def qaly(self) -> float: ...
+    @property
+    def se(self) -> float: ...
+    @property
+    def lower(self) -> float: ...
+    @property
+    def upper(self) -> float: ...
+
+class RCLLResult:
+    @property
+    def rcll(self) -> float: ...
+    @property
+    def se(self) -> float: ...
+    @property
+    def n_samples(self) -> int: ...
+
+def survfitkm(
+    time: list[float],
+    status: list[int],
+    conf_level: float = 0.95,
+) -> SurvFitKMOutput: ...
+def survfitkm_with_options(
+    time: list[float],
+    status: list[int],
+    options: SurvfitKMOptions,
+) -> SurvFitKMOutput: ...
+def nelson_aalen_estimator(
+    time: list[float],
+    status: list[int],
+    conf_level: float = 0.95,
+) -> NelsonAalenResult: ...
+def stratified_kaplan_meier(
+    time: list[float],
+    status: list[int],
+    strata: list[int],
+    conf_level: float = 0.95,
+) -> dict[int, SurvFitKMOutput]: ...
+def logrank_test(
+    time: list[float],
+    status: list[int],
+    group: list[int],
+) -> LogRankResult: ...
+def fleming_harrington_test(
+    time: list[float],
+    status: list[int],
+    group: list[int],
+    rho: float = 0.0,
+    gamma: float = 0.0,
+) -> LogRankResult: ...
+def rmst(
+    time: list[float],
+    status: list[int],
+    tau: float,
+    conf_level: float = 0.95,
+) -> RMSTResult: ...
+def rmst_comparison(
+    time: list[float],
+    status: list[int],
+    group: list[int],
+    tau: float,
+    conf_level: float = 0.95,
+) -> dict[str, Any]: ...
+def survreg(
+    time: list[float],
+    status: list[int],
+    x: list[list[float]],
+    config: SurvregConfig | None = None,
+) -> SurvivalFit: ...
+def calibration(
+    predicted: list[float],
+    time: list[float],
+    status: list[int],
+    n_groups: int = 10,
+) -> CalibrationResult: ...
+def uno_c_index(
+    time: list[float],
+    status: list[int],
+    risk_scores: list[float],
+    tau: float | None = None,
+) -> UnoCIndexResult: ...
+def brier(
+    time: list[float],
+    status: list[int],
+    predicted: list[float],
+    eval_time: float,
+) -> float: ...
+def integrated_brier(
+    time: list[float],
+    status: list[int],
+    predicted: list[list[float]],
+    eval_times: list[float],
+) -> float: ...
+def bootstrap_cox_ci(
+    time: list[float],
+    status: list[int],
+    x: list[list[float]],
+    n_bootstrap: int = 1000,
+    conf_level: float = 0.95,
+) -> BootstrapResult: ...
+def cv_cox_concordance(
+    time: list[float],
+    status: list[int],
+    x: list[list[float]],
+    n_folds: int = 5,
+) -> CVResult: ...
+def lrt_test(
+    log_likelihood_full: float,
+    log_likelihood_reduced: float,
+    df: int,
+) -> TestResult: ...
+def bayesian_cox(
+    time: list[float],
+    status: list[int],
+    x: list[list[float]],
+    n_samples: int = 1000,
+    n_warmup: int = 500,
+    prior_scale: float = 1.0,
+) -> BayesianCoxResult: ...
+def elastic_net_cox(
+    time: list[float],
+    status: list[int],
+    x: list[list[float]],
+    alpha: float = 1.0,
+    l1_ratio: float = 0.5,
+    max_iter: int = 1000,
+) -> ElasticNetCoxResult: ...
+def mixture_cure_model(
+    time: list[float],
+    status: list[int],
+    x_cure: list[list[float]],
+    x_survival: list[list[float]],
+    max_iter: int = 100,
+) -> MixtureCureResult: ...
+def joint_model(
+    longitudinal_time: list[float],
+    longitudinal_value: list[float],
+    subject_id: list[int],
+    survival_time: list[float],
+    survival_status: list[int],
+    x: list[list[float]],
+) -> JointModelResult: ...
+def ipcw_kaplan_meier(
+    time: list[float],
+    status: list[int],
+    weights: list[float],
+) -> SurvFitKMOutput: ...
+def g_computation(
+    time: list[float],
+    status: list[int],
+    treatment: list[int],
+    x: list[list[float]],
+    n_bootstrap: int = 1000,
+) -> GComputationResult: ...
+def qaly_calculation(
+    time: list[float],
+    status: list[int],
+    utility: list[float],
+    tau: float,
+) -> QALYResult: ...
+def rcll(
+    time: list[float],
+    status: list[int],
+    predicted_survival: list[list[float]],
+    eval_times: list[float],
+) -> RCLLResult: ...
+def gradient_boost_survival(
+    x: list[float],
+    n: int,
+    p: int,
+    time: list[float],
+    status: list[int],
+    config: GradientBoostSurvivalConfig | None = None,
+) -> GradientBoostSurvival: ...
+def survival_forest(
+    x: list[float],
+    n: int,
+    p: int,
+    time: list[float],
+    status: list[int],
+    config: SurvivalForestConfig | None = None,
+) -> SurvivalForest: ...
+def load_lung() -> dict[str, list[Any]]: ...
+def load_aml() -> dict[str, list[Any]]: ...
+def load_veteran() -> dict[str, list[Any]]: ...
+def load_ovarian() -> dict[str, list[Any]]: ...
+def load_colon() -> dict[str, list[Any]]: ...
+def load_pbc() -> dict[str, list[Any]]: ...
+def load_cgd() -> dict[str, list[Any]]: ...
+def load_bladder() -> dict[str, list[Any]]: ...
+def load_heart() -> dict[str, list[Any]]: ...
+def load_kidney() -> dict[str, list[Any]]: ...
+def load_rats() -> dict[str, list[Any]]: ...
+def load_stanford2() -> dict[str, list[Any]]: ...
+def load_udca() -> dict[str, list[Any]]: ...
+def load_myeloid() -> dict[str, list[Any]]: ...
+def load_flchain() -> dict[str, list[Any]]: ...
+def load_transplant() -> dict[str, list[Any]]: ...
+def load_mgus() -> dict[str, list[Any]]: ...
+def load_mgus2() -> dict[str, list[Any]]: ...
+def load_diabetic() -> dict[str, list[Any]]: ...
+def load_retinopathy() -> dict[str, list[Any]]: ...
+def load_gbsg() -> dict[str, list[Any]]: ...
+def load_rotterdam() -> dict[str, list[Any]]: ...
+def load_logan() -> dict[str, list[Any]]: ...
+def load_nwtco() -> dict[str, list[Any]]: ...
+def load_solder() -> dict[str, list[Any]]: ...
+def load_tobin() -> dict[str, list[Any]]: ...
+def load_rats2() -> dict[str, list[Any]]: ...
+def load_nafld() -> dict[str, list[Any]]: ...
+def load_cgd0() -> dict[str, list[Any]]: ...
+def load_pbcseq() -> dict[str, list[Any]]: ...
+def load_hoel() -> dict[str, list[Any]]: ...
+def load_myeloma() -> dict[str, list[Any]]: ...
+def load_rhdnase() -> dict[str, list[Any]]: ...