Main Page

Welcome to the Collective Computational Unit (CCU)

The CCU handles computer science and infrastructure aspects for the Centre of the Advanced Study of Collective Behaviour. It constitutes a core facility that is responsible for providing technical platforms and support, algorithmic solutions, and software frameworks, and will be home to technicians and a number of funded student assistants as well as computer science students at all levels. As a collaborative research enterprise, we provide and develop algorithms for the analysis and visualisation of highly complex scientific data repositories. In particular, the focus is on efficient methods for tracking and measuring collectives, (semi) automated data analysis, data mining, numerical simulation and parameter estimation, and visual and interactive data exploration.

The CCU offers three types of project support:

• Standard Problems can be approached using well-established software tools (such as Amira, MatLab, KNIME). The centre will provide help with using software packages to solve problems and offer additional expertise, if required. This help will be provided primarily by student research assistants.
• Elaborate Problems require modifications to or plugins for software packages involving known algorithms. These problems will be handled by master’s students specialising in computer science, who will either be paid for the support they provide, or, if the problem is more scientific rather than just technical in nature, will be given the opportunity to work on it as part of their master’s thesis.
• Special Problems cannot be solved using standard approaches or software. To address these kinds of problems, the centre will provide help with finding the right experts in computer science or mathematics and also offer collaboration in our own fields of expertise. Projects in need of more comprehensive and complex algorithmic solutions that exceed initial estimates will have to apply for doctoral or post-doctoral researchers via the centre’s open calls for projects. These researchers will receive support from the computational centre in the form of regular meetings in which algorithmic techniques will be discussed.

About this Wiki

This Mediawiki is intended to link the many challenging problems arising in data analysis for collectives to experts from computer science, and provide a global platform for discussion to which everyone can contribute. Over time, it will also become a knowledge base and reference for methods and expertise available in the CCU. In particular, it introduces and gives tutorials for using computational resources, code repositories and data storage for all members.

Remember a Wiki lives from the contributions of all members. In particular, please make ample use of the discussion page which is available for every individual page (see "Discussion" link on top for an example), and leave comments whenever you think you have something useful to add.

Important links

• Current roadmap for the CCU
• List of open/ongoing projects
• Suggestions for improving services
• Bugs and problems

Getting started

• General

- Becoming a registered user

- Accessing the GPU cluster

• Projects

- Adding a new open project

- Tracking progress of a project

- Contributing to a project

- Closing a project

• Computational resources

- Accessing the docker container repository

- Wrapping code in a container to run on the cluster

- Running jobs on the cluster

- Uploading data to the global storage

- Accessing data (for network training etc.) from the GPU cluster

- Using our git server to maintain your code

• Mediawiki

- Consult the Mediawiki User's Guide for information on using the wiki software.

Contact

Primary CCU contact: | Bastian Goldluecke