API Reference

This is the class and function reference of chainladder. Please refer to the full user guide for further details, as the class and function raw specifications may not be enough to give full guidelines on their uses.

chainladder.core: Triangle

Classes

Triangle([data, origin, development, ...])	The core data structure of the chainladder package
DevelopmentCorrelation(triangle[, p_critical])	Mack (1997) test for correlations between subsequent development factors.
ValuationCorrelation(triangle[, p_critical, ...])	Mack (1997) test for calendar year effect.A calendar period has impact across developments if the probability of the number of small (or large) development factors, Z, in that period occurring randomly is less than p_critical

chainladder.development: Development Patterns

Classes

Development([n_periods, average, ...])	A Transformer that allows for basic loss development pattern selection.
DevelopmentConstant([patterns, style, ...])	A Estimator that allows for including of external patterns into a
MunichAdjustment([paid_to_incurred, fillna])	Applies the Munich Chainladder adjustment to a set of paid/incurred
IncrementalAdditive([trend, n_periods, ...])	The Incremental Additive Method.
ClarkLDF([growth, groupby])	A Estimator that allows for curve fitting development pattterns according to Clark 2003.
CaseOutstanding([paid_to_incurred, ...])	A determinisic method based on outstanding case reserves.
TweedieGLM([design_matrix, response, ...])	This estimator creates development patterns with a GLM using a Tweedie distribution.
DevelopmentML([estimator_ml, y_ml, ...])	A Estimator that interfaces with machine learning (ML) tools that implement the scikit-learn API.
BarnettZehnwirth([formula, response])	This estimator enables modeling from the Probabilistic Trend Family as described by Barnett and Zehnwirth.

chainladder.tails: Tail Factors

tails should store all tail methodologies

Classes

TailConstant([tail, decay, attachment_age, ...])	Allows for the entry of a constant tail factor to LDFs.
TailCurve([curve, fit_period, ...])	Allows for extraploation of LDFs to form a tail factor.
TailBondy([earliest_age, attachment_age, ...])	Estimator for the Generalized Bondy tail factor.
TailClark([growth, truncation_age, ...])	Allows for extraploation of LDFs to form a tail factor.

chainladder.methods: IBNR Methods

Classes

Chainladder()	The basic determinsitic chainladder method.
MackChainladder()	Basic stochastic chainladder method popularized by Thomas Mack
BornhuetterFerguson([apriori, ...])	The deterministic Bornhuetter Ferguson IBNR model
Benktander([apriori, n_iters, ...])	The Benktander (or iterated Bornhuetter-Ferguson) IBNR model
CapeCod([trend, decay, n_iters, ...])	Applies the CapeCod technique to triangle X

chainladder.workflow: Adjustments

Classes

BootstrapODPSample([n_sims, n_periods, ...])	Class to generate bootstrap samples of triangles.
BerquistSherman([paid_amount, ...])	Class to alter the inner diagonals of a Triangle using the methods described by Berquist and Sherman.
Trend([trends, dates, axis])	Estimator to create and apply trend factors to a Triangle object.
ParallelogramOLF([rate_history, change_col, ...])	Estimator to create and apply on-level factors to a Triangle object.

chainladder.workflow: Workflow

Classes

Pipeline(steps, *[, memory, verbose])	This is a near direct of copy the scikit-learn Pipeline class.
VotingChainladder(estimators, *[, weights, ...])	Prediction voting chainladder method for unfitted estimators.
GridSearch(estimator, param_grid, scoring[, ...])	Exhaustive search over specified parameter values for an estimator.

chainladder.utils: Utilities

utils should store all utility functions and classes, i.e. things that are used by various modules in the package.

Functions

load_sample(key, *args, **kwargs)	Function to load datasets included in the chainladder package.
read_pickle(path)
read_json(json_str[, array_backend])
concat(objs, axis[, ignore_index, sort])	Concatenate Triangle objects along a particular axis.
load_sample(key, *args, **kwargs)	Function to load datasets included in the chainladder package.
minimum(x1, x2)
maximum(x1, x2)

Classes

PatsyFormula([formula])

A sklearn-style Transformer for patsy formulas.