InnocentMoTMo

Scripts for the analysis of the innocent agent rates from the MoTMo model. Application of the Augmented Jump Chain (AJC), PCCA+ and the shortest path problem to extract the most likely transition path between metastable clusters in a nonautonomous setting.

Contents

• clusterQtxt.jl: read the sparse single timestep matrix ratesStep01.txt and compute the Schur/PCCA $\chi$ function
• plot.jl: Visualization of the 4-dimensional $\chi$ function on the 3-simplex / tetrahedron.
• shortestpath.jl: Compute the most likely (shortest in inverse rates) path between two clusters / vertices.
• presentation.jl: Reconstruct the plot from the presentation (transition paths of the AJC).

Installation

• Besides the packages in the Project.toml it requires the non-published PCCA.jl:
  using Pkg; Pkg.add("https://github.com/axsk/PCCA.jl")
• Also, depending on the usage it requires the whole AJC matrix Q_matrix.mat and chi4.mat or the single timestep rate matrix ratesStep01.txt.