Algorithms and computation

Diplome(s)
Lieu
Université Paris Cité
Automne - Hiver
Niveau Master 2 6 ECTS - En anglais
Enseignant(s) Ludovic Berthier ( CNRS ) Michel Ferrero ( Polytechnique )
Contact - Secrétariat de l’enseignement

Tél : + 33 (1) 44 32 35 60 
enseignement@phys.ens.fr


Contact - Secrétariat pédagogique

Tel : + 33 (1) 44 27 40 70 
Evelyne Gilbert Mongeot


Correspondants

Sorbonne Université 
Jerome.Tignon@phys.ens.fr 

Université Paris Cité 
edouard.boulat@univ-paris-diderot.fr 

Université Paris-Saclay 
pascal.simon@u-psud.fr 

Institut Polytechnique de Paris 
silke.biermann@polytechnique.fr 

Ecole Normale Supérieure 
giuli.biroli@ens.fr

Computational physics plays a central role in all fields of physics, from classical statistical physics, soft matter problems, and hard-condensed matter. Our goal is to cover the basic concepts underlying computer simulations in classical and quantum problems, and connect these ideas to relevant and contemporary research topics in various fields of physics. In the TD’s you will also learn how to set, perform and analyse the results of simple computer simulations by yourself, covering a wide range of topics. We will use Python, but no previous knowledge of this programming language is needed.

Syllabus

1) Introduction to computational physics 
Image retirée. Brief history and examples

2) Monte Carlo methods 
Image retirée. Markov chains, Metropolis algorithm 
Image retirée. Statistical analysis of errors

3) Monte Carlo simulations of Ising models 
Image retirée. Application to Ising models and non-equilibrium phenomena

4) Monte Carlo simulations for off-latice models 
Image retirée. Some interesting details : Forces, neighbors, hard spheres 
Image retirée. Application to classical fluids

5) Molecular Dynamics 
Image retirée. Equations of motion, statistical ensembles

6) Complex and disordered systems 
Image retirée. Interesting problems and solutions : simulated annealing, parallel tempering

7) More complex situations 
Image retirée. Umbrella sampling, reweighting and phase transitions studies

8) Matrix diagonalisation in quantum physics 
Image retirée. Lanczos or Density Matrix Renormalisation Group for quantum spin systems

9) Path integral approach to quantum problems 
Image retirée. Quantum particles in arbitrary potentials

Prerequisites

The lectures cover a broad range of topics, and basic knowledge of statistical and quantum physics is required. Connections are made to contemporary research topics. 

Regarding computer and simulations themselves, the tutorials are an ideal place to learn how to use and implement python, and how to organise the simulations in a python notebook, but no previous knowledge of python is needed to be able to participate. 

 

Evaluation

The evaluation is based on two steps: a homework exploring slightly more advanced topics compared to the tutorials will be given in the middle of the lectures. At the end of the lectures a formal exam will take place, with a fixed limited time with both questions regarding the lectures and practical ones related to the tutorials. The overall goal of the evaluation is to check whether the concepts, results, and tools developed in the tutorials have been understood.