Loglikelihood Adjustment for Extreme Value Models

Fits the mixed cumulative incidence function models suggested by Cederkvist et al (2018; https://doi.org/10.1093%2Fbiostatistics%2Fkxx072) which decomposes within cluster dependence of risk and timing.

Code for Ren, B. and Barnett, I. (2022), Autoregressive mixture models for clustering time series. J. Time Ser. Anal., 43: 918-937. https://doi.org/10.1111/jtsa.12644 https://onlinelibrary.wiley.com/doi/abs/10.1111/jtsa.12644