This repository is a collection of code written in R used to analyse the price behaviour and volatility of Bitcoin prior to the conclusion of a bubble-like phase. The code was initially used for my university thesis but can be adapted to fit any price series.
It uses the Log-Periodic-Power-Law (LPPL) model proposed by Sornette, Johansen et al (1998, 2000) and fits it according to i) a model transformation (see Filiminov and Sornette, 2013); followed by ii) an evolutionary search algorithm (see Jacobsson, 2009). Also included are diagnostic tests, model checks and sensitivity analysis.