Reproducible and Interactive Data Analysis and Modelling using Jupyter Notebooks (study period II, 4 ECTS)
The aim of this course is to introduce students to the Jupyter Notebook which is an open-source web application that allows you to create and share documents that contain live code, equations, visualizations and explanatory text. Uses include: data cleaning and transformation, numerical simulation, statistical modeling, machine learning and much more. Through the notebooks, research results and the underlying analysis can be transparently reproduced as well as shared.
During three days with alternating lectures and hands-on exercises, the participants will learn to construct well-documented, electronic notebooks that perform advanced data analysis and produce publication ready plots. While the course is based on Python, this is not a prerequisite, and many other programming languages can be used.
For more details see the course website.
Course teachers: , ,
Domain Decomposition Methods (study period I, 7.5 ECTS)
Domain Decomposition is an important concept in the parallelization of numerical methods for partial differential equations (PDEs). Thereby, the spatial domain is decomposed into several parts, which are assigned to different processes, which then communicate over the boundaries of these domains. Naturally the question arises what happens if we modify the data or reduce the amount of communication. For large classes of PDEs, practical answers can be found by mathematical analysis, which turns out to be relevant to other multiphysics problems such as fluid structure interaction as well.
The course will cover the basic theory of domain decomposition methods for model problems like the Poisson and the Heat Equation with finite difference discretizations and then look at the practical issues of implementing these in a Message Passing Interface (MPI) framework.
Course responsible: ,
Distributed computing concepts and tools (study period II, 4.5 ECTS)
The course gives introduction into concepts of geographically distributed computing, such as that implemented by research e-infrastructures dealing with large data volumes and high processing rates. The course is a combination of lectures and hands-on tutorials, addressing aspects of security, authentication and authorisation, interfaces to computing services, principles of distributed storage and data handling, resource characterisation and discovery, information representation and monitoring, workload management and scheduling in a distributed environment. It introduces concepts of services, interfaces, resource virtualisation, meta-protocols, non-interactive workloads, digital certificates and trust, Virtual Organisations, information systems and execution environments. Existing tools and services are introduced as well. The students will obtain personal certificates and a temporary access to a distributed computing infrastructure for exercises. Assessment is done on the basis of the course project, during which the students are expected to make use of distributed computing to solve their daily tasks.
Teachers: Oxana Smirnova, Balazs Konya, Florido Paganelli
Course responsible: Oxana Smirnova
Members of COMPUTE will be notified by e-mail when each course is open for
registration. For further questions please contact the study directors:
List of previous courses here.
This page was last modified on 20 November 2017.