Training - Extending ParaView

Scientific visualization is the art of utilizing human visual cognition as a means to glean understanding and insight into scientific data, be it computed numerically or acquired by measurement.

Together with data analysis, it is a key component of today's scientific discovery process, in particular in the context of computational sciences, with application ranging from astrophysics to climatology, and from electromagnetics to mechanical engineering.​

The visualization toolkit (VTK) is an open-source, C++ software library dedicated to data visualization, image processing and computer graphics. Also offering Python and Java interfaces, it supports a broad class of visualization technique for a variety of data and attributes types.

Used extensively by thousands of users worldwide, it also constitutes the visualization engine of ParaView, an also open-source, multi-platform, application designed to visualize and analyse data on a broad set of platforms, ranging from on single processor on a laptop, to many thousands of cores on a high-performance computing cluster.

Both VTK and ParaView have achieved global recognition in the scientific visualization community over the past fifteen years. ParaView in particular has made major inroads into scientific research of today, as datasets have been growing at an extremely rapid pace and are continuing to do so.

The goal of this course is to offer to users of paraView an extended knowledge in terms of customization, extension, and parallel usage.
The trainee will learn how to use VTK features that are not exposed in the ParaView's user interface, as well as to design and develop fully new capabilities.

Depending on the customer's request, the course will provide several options: either an introduction to parallel data visualization with ParaView or the development of C++ plugins, or both (in which case the duration of the course means an additional half of a day).

This course, taught by a computational scientist and developer with over a decade worth of experience with both VTK and ParaView is intended for an audience of computational as well as application scientists, engineers, post-docs, and students with a background or an interest in numerical simulation.

Pre-requisites

• Basic knowledge of VTK
• Good knowledge of Python
• Some knowledge of C++ is required for the "ParaView Plugins" option
• Some knowledge of MPI is useful for the "parallel ParaView" option

PART 1

• Customizing ParaView (about 3 hours)
• Custom and programmable filters, server-side Python scripting

PART 2

• CParaView scripting (about 3 hours)
• Client-side Python scripting

PART 3 and/or PART 4

• Extending ParaView with C++ plugins or parallel ParaView or both (to be decided prior to the course)

Duration: 1 day (or 1 day and a half if both options of parallel ParaView and ParaView plugins are chosen)

Language: English or French

Location: At the client's premises

Lecturer: Dr Philippe Pébaÿ