UniLindleyApprox News
Version 0.1.0 (2024-07-19)
Initial Release
- Implemented core Lindley approximation engine for Bayesian point
estimation
- Added support for 23 censoring and truncation schemes
- Implemented 8 loss functions for Bayes estimation
- Added automatic numerical derivative computation (first, second,
third-order)
- Implemented comprehensive S3 methods for lindleyfit objects
- Added goodness-of-fit statistics (KS, AD, CvM, Watson,
Chi-square)
- Implemented residual analysis (Cox-Snell, Martingale, Deviance,
Pearson, Generalized, Randomized quantile)
- Added visualization functions (posterior surfaces, likelihood
profiles, QQ/PP plots, residual plots)
- Implemented prediction utilities (quantiles, survival, hazard,
cumulative hazard)
- Added simulation utilities for all censoring schemes
- Implemented model selection criteria (AIC, AICc, BIC, HQIC, CAIC,
KIC)
- Added support for multiple prior distributions (Gamma, Beta, Normal,
Uniform, Lognormal, Inverse Gamma, Weibull, Exponential, Jeffreys)
- Implemented modular, extensible architecture for adding new
censoring schemes
- Added comprehensive error handling and validation
- Created roxygen2 documentation for all functions