chainladder-python
chainladder-python copied to clipboard
casact
Actuarial reserving in Python
Many assumptions/hypterparameters of Estimators take on a vector-like feel along an axis of a Triangle. However, there is a great deal of inconsistency in implementation across estimators with at least...
Need to create cross-validation functionality that holds out valuation dates similar to `sklearn.model_selection.TimeSeriesSplit`.
Prototype implementation of [Clark LDF Loglogistic curve fit](https://www.casact.org/pubs/forum/03fforum/03ff041.pdf): ```python import chainladder as cl import numpy as np from scipy.optimize import minimize self = cl.load_sample('genins') # Put data in tablular form...
Add method to do Baesian MCMC. See: [CAS Monograph #1](https://www.casact.org/pubs/monographs/index.cfm?fa=meyers-monograph01)
Need to include Shapland's hetero grouping option in ODP Bootstrap simulation.
The [current version](https://github.com/casact/chainladder-python/blob/ba303e4fdc72ec646afbb8013f932a67a756b870/chainladder/core/correlation.py#L277) of the valuation correlation algorithm uses the variance to construct the confidence interval, and then compares the variance to the two endpoints of the confidence interval: ```...
I have some proposals for changing the output formatting of the three Mack tests from the first part of the development tutorial: 1. Valuation Correlation - All Origins: Currently output...
While looking into #320, I noticed you can input values lower than 0 and higher than 1 for the critical value in the Mack valuation correlation test, without raising an...
drop_high does some weird stuff sometimes ``` prism_df = pd.read_csv( "https://raw.githubusercontent.com/casact/chainladder-python/master/chainladder/utils/data/prism.csv" ) prism_df["AccidentDate"] = pd.to_datetime(prism_df["AccidentDate"]) prism_df["PaymentDate"] = pd.to_datetime(prism_df["PaymentDate"]) prism_df.loc[(prism_df["AccidentDate"].dt.year > 2008) & (prism_df["PaymentDate"].dt.year > 2012),"Incurred"] = 0 prism = cl.Triangle(...
Prolly something dumb I'm doing. This code works. ``` raa = cl.load_sample('raa') model = cl.Development().fit(raa) model.std_residuals_ ``` This code doesn't. ``` raa = cl.load_sample('raa') model = cl.Development().fit_transform(raa) model.std_residuals_ ```