Elsevier Science Home
Computer Physics Communications Program Library
Full text online from Science Direct
Programs in Physics & Physical Chemistry
CPC Home

[Licence| Download | New Version Template] afao_v1_0.tar.gz(2726 Kbytes)
Manuscript Title: FabSim: facilitating computational research through automation on large-scale and distributed e-infrastructures
Authors: Derek Groen, Agastya Bhati, James Suter, James Hetherington, Stefan Zasada, Peter V. Coveney
Program title: FabSim
Catalogue identifier: AFAO_v1_0
Distribution format: tar.gz
Journal reference: Comput. Phys. Commun. 207(2016)375
Programming language: Python.
Computer: PC or Mac.
Operating system: Unix, OSX.
RAM: 1 Gbytes
Keywords: Automation, Workflows, Distributed computing, Software, Bloodflow modelling, Molecular dynamics, Lattice-Boltzmann, Multiscale modelling, Clay- polymer nanocomposites, Ligand-protein binding.
Classification: 3, 4, 6.5.

External routines: NumPy, SciPy, Fabric (1.5 or newer), PyYaml

Nature of problem:
Running advanced computations using remote resources is an activity that requires considerable time and human attention. These activities, such as organizing data, configuring software and setting up individual runs often vary slightly each time they are performed. To lighten this burden, we required an approach that introduced little burden of its own to set up and adapt, beyond which very substantial productivity ensues.

Solution method:
We present a toolkit which helps to simplify and automate the activities which surround computational science research. FabSim is aimed squarely at the experienced computational scientist, who can use the command line interface and a system of modifiable content to quickly automate sets of research tasks.

Restrictions:
FabSim relies on a command-line interface, and assumes some level of scripting knowledge from the user.

Unusual features:
FabSim has a proven track record of being easy to adapt. It has already been extensively adapted to facilitate leading research in the modelling of bloodflow, nanomaterials, and ligand-protein binding.

Running time:
FabSim can be used interactively, typically requiring a few seconds to perform a basic task.