Sampling high-dimension probability measures: applications in (non)equilibrium molecular dynamics and statistics
This repository contains the various files for the hands-on sessions of the SINEQ summer school @ CERMICS, September 2023. The schedule can be found here. The files for each sessions are given below.
- Monday, Sept. 25: Discretization of Langevin dynamics and its Metropolization
- Tuesday, Sept. 26: Nonequilibrium systems and coupling methods
- Wednesday, Sept. 27, morning: Accelerated molecular dynamics methods
- Wednesday, Sept. 27, afternoon: Adaptive multilevel splitting methods (Answers)
- Thursday, Sept. 28: Computing average properties in molecular dynamics with Molly.jl
- Friday, Sept. 29: Computation of transport coefficients with Molly.jl
We provide Julia notebooks in this repository. The three main ingredients required are Python (>=3.3 or 2.7, needed for Jupyter Notebook), Jupyter Notebook and Julia.
-
If Python is not installed on your system, install it using Anaconda. It conveniently installs Python, the Jupyter Notebook, and other commonly used packages for scientific computing and data sciences. You can download Anaconda here (install it with the latest Python version). Then install the version of Anaconda which you downloaded, following the instructions on the download page. Once installed, you can simply launch Jupyter Notebook by running in a shell:
jupyter-notebook
-
If you already have Python installed (>= 3.3 or 2.7), run the following commands in a shell (use pip instead of pip3 if you use 2.7):
pip3 install --upgrade pip pip3 install jupyter
Please refer to https://github.com/JuliaLang/juliaup where you will find all the needed resource to install Julia on your OS.
Once installed, open a Julia REPL by typing julia
in a shell. Then type:
using Pkg
Pkg.add("IJulia")
so that Julia can be run inside Jupyter Notebooks.
Run the notebook installation.ipynb
using the following in a shell:
jupyter-notebook installation.ipynb
For the Molly session, the molecular visualization software PyMol is recommended. You can install it by following the instructions provided here.