Algorithm implementations and case studies for the paper Accelerating delayed-acceptance Markov chain Monte Carlo algorithms

This repository contains all code for the pre-print paper Accelerating delayed-acceptance Markov chain Monte Carlo algorithms https://arxiv.org/abs/1806.05982.

N.B.: The results in the pre-print at arXiv v2 are computed using the version of the code at tag preprint_v2.

All computations are carried out on the AURORA@LUNARC http://www.lunarc.lu.se/resources/hardware/aurora/ cluster.

The lunarc branch is deprecated, only the master branch is used now.

File structure

The files are structured as following:

/DWPSDE model

• source files for the implementations of the algorithms for the DWP-SDE model
• scripts to run the algorithms
• run-scripts for LUNARC
• datasets and simulated data
• training data for DA and ADA

/DWPSDE model/Results

• all numerical results in .csv files

/DWPSDE model/lunarc_output

• output files from LUNARC

/DWPSDE model/analyses

• scripts for analyzing the results

/Ricker model

• source files for the implementations of the algorithms for the Ricker model
• scripts to run the algorithms
• simulated data
• training data for DA and ADA
• run-scripts for LUNARC
• scripts for analyzing the results

/Ricker model/Results

• all numerical results in .csv files The protein folding data is provided in the files:

DWPSDE model/new_data_set.txt DWPSDE model/1LE1_L.dat

/Ricker model/lunarc_output

• output files from LUNARC

/adaptive update algorithms

• source files for the adaptive tuning algorithms

/gpmodel

• source files for the GP model

/select cases

• source files for the different methods to select which case to assume in the ADA algorithm

/utilities

• source files for various help functions

Software

Julia version

julia> versioninfo()
Julia Version 0.5.2
Platform Info:
  OS: Linux (x86_64-pc-linux-gnu)
  CPU: Intel(R) Xeon(R) CPU E5-2650 v3 @ 2.30GHz
  WORD_SIZE: 64
  BLAS: libmkl_rt
  LAPACK: libmkl_rt
  LIBM: libimf
  LLVM: libLLVM-3.7.1 (ORCJIT, haswell)

Packages

julia> include("utilities/print_used_packages.jl")

 - Distributions                 0.15.0
 - DataFrames                    0.10.1
 - StatsBase                     0.19.2
 - Optim                         0.7.8
 - Lasso                         0.1.0
 - StatsFuns                     0.5.0
 - JLD                           0.8.3
 - HDF5                          0.8.8
 - PyPlot                        2.3.2+             master
 - KernelDensity                 0.4.0

Data

The protein folding data is provided in the files:

DWPSDE model/new_data_set.txt

DWPSDE model/1LE1_L.dat