Result of ``.fit()`` or ``.from_params()`` method for every parametric surpyval distribution. Instances of this class are very useful wh

Normalise and hold row-aligned multivariate survival data. Parameters ---------- x : array-like, shape (N, D) or sequence of D length-N a

Result of ``.fit()`` or ``.from_params()`` method for parametric regression modelling. Instances of this class are very useful when a us

A fitted renewal / imperfect-repair recurrence model. This is the model object returned by the renewal-family fitters (``GeneralizedRene

Pure-NumPy grouped aggregation, replacing numpy_indexed.group_by.

A class for holding the parameters, data, and usefult methods for a fitted parametric recurrence model. This is the result of the ``fit`` cal

A fitted bivariate copula glued to two univariate margins. Attributes ---------- copula : Copula The copula family. params :

Result of ``.fit()`` method for every non-parametric surpyval distribution. This means that each of the methods in this class can be call

r""" Class used to generate the Normal (Gauss) class. .. code:: python from surpyval import Normal

Model to provide methods and attributes when using a fitted proportional intensity model. Simulation reuses the shared :class:`Recurrenc

Result of a recurrent-event trend test (:func:`laplace` or :func:`mil_hdbk_189c`). Attributes ---------- statistic : float

Accelerated Failure Time fitter using exp(beta'Z) as the acceleration factor. The cumulative hazard is: H(x | Z) = H_0(exp(beta'

r""" The four-parameter (generalised) Beta distribution. The standard :class:`Beta` distribution is supported on ``[0, 1]``. The four-par

r""" The Binomial distribution: the number of events (failures) ``k`` in a fixed number ``n`` of independent pass/fail trials, each with event

Clayton copula (lower-tail dependence), ``theta > 0``.

Abstract base class for parametric counting-process intensity models. A counting process ``N(t)`` counts the number of events observed up to

Dual Exponential Life Model This class represents a dual exponential life model, which is used for survival analysis. Attributes

r""" Class used to generate the Exponential class. .. code:: python from surpyval import Exponential

r""" Fleming-Harrington estimation of survival distribution. Returns a `NonParametric` object from method :code:`fit()` calculates the Non

Frank copula (symmetric, no tail dependence), ``theta != 0``.

r""" Class used to generate the Gamma class. .. code:: python from surpyval import Gamma

Gaussian copula, ``rho in (-1, 1)`` (no tail dependence).

Gumbel-Hougaard copula (upper-tail dependence), ``theta >= 1``.

The independence copula ``C(u, v) = u v`` (no parameter).

Contract shared by the closed-form NHPP intensity baselines (:class:`Crow-AMSAA <surpyval.recurrent.parametric.crow_amsaa.CrowAMSAA_>`, :

r""" Kaplan-Meier estimator class. Calculates the Non-Parametric estimate of the survival function using: .. math:: R(x) = \prod_

Result of a (weighted) log-rank test. Attributes ---------- statistic : float The chi-squared test statistic. dof : int

r""" Nelson-Aalen estimator class. Returns a `NonParametric` object from method :code:`fit()` Calculates the Non-Parametric estimate of th

Proportional Odds model fitter using exp(beta'Z) as the odds multiplier. The survival odds satisfy: O(x | Z) = O_0(x) * exp(beta'Z)

r""" Turnbull estimator class. Returns a `NonParametric` object from method :code:`fit()`. Calculates the Non-Parametric estimate of the survi

Internal phi object: phi(Z) = exp(beta'Z).

Arithmetic Reduction of Age (ARA) imperfect-repair model of Doyen and Gaudoin (2004). Each repair removes a fraction of the accumulated

Arithmetic Reduction of Intensity (ARI) imperfect-repair model of Doyen and Gaudoin (2004). Where the ARA/Kijima models reduce the *virt

Cause-specific Mean Cumulative Function for a recurrent process with competing event types. The model fits one ``NonParametricCounting``

TODO: Time-Varying Implementation TODO: Change this to SemiParametricCompetingRiskProportionalHazard ??

Bivariate copula family. Subclasses define :meth:`cdf` (and, for speed/stability, may override the partial derivatives, dependence measures a

A class to represent the Cox-Lewis model for non-homogeneous Poisson processes (NHPP). This model is used in reliability analysis to predict

A class to represent the Crow-AMSAA model for non-homogeneous Poisson processes (NHPP). This model is used in reliability analysis to predict

Used to create a custom distribution using only the cumulative hazard function. The cumulative hazard function must be a function of x and

Mixin adding a ``fit_from_df`` method to a parametric regression fitter. The fitter must expose a ``fit(x, Z, c=None, n=None, t=None, init=N

Root abstract base class that every surpyval model inherits from. The contract shared by all models -- parametric, nonparametric, mixtur

Represents the Duane Non-Homogeneous Poisson Process model. This class includes methods to evaluate various statistical functions of the

A class to handle the G1 renewal process of Kaminskiy and Krivtsov, in which the jth interarrival time is the underlying lifetime distributio

A class to handle the generalized renewal process with different Kijima models. Since the Generalised Renewal Process does not have clos

Represents the Homogeneous Poisson Process (HPP) model. This class includes methods to evaluate various statistical functions of the mode

Likelihood-based inference for fitted recurrent-event models. The fitting routine must set ``_neg_ll`` (the negative log-likelihood in n

A class for creating a Mixture Model fitter. This class implements a Mixture Model, which is a probabilistic model that combines multipl

A jointly-specified model of several correlated event-time series (the ``multivariate`` outcome-dimension axis of ``MODEL_ATLAS.md``). U

The common methods between IntermediateNode and LeafNode.

An empirical model produced by a nonparametric estimator (Kaplan-Meier, Nelson-Aalen, Fleming-Harrington or Turnbull). Adds random sampli

A fully specified parametric model. In addition to the survival interface it supports random sampling and the standard statistical summar

Base class for all parametric distributions. A distribution needs only ``hf`` and ``Hf`` (or ``sf``, ``ff`` and ``df``) plus a ``_parame

A class representing the Proportional Intensity Homogeneous Poisson Process (HPP). The class contains methods to perform various calcula

A class representing the Proportional Intensity Non-Homogeneous Poisson Process (NHPP). The class contains methods to perform various ca

Random Survival Forest Specs: - n_trees `Tree`'s trained, each given independently bootstrapped samples - Each tree is trained - Each

Shared simulation machinery for fitted recurrent-event models. Subclasses provide the per-event sampling logic by implementing ``_new_se