Tulip is an open source, cross-platform, data visualization framework mainly dedicated to the analysis, the drawing and the visualization of very large graphs (up to the million of nodes and edges). It has been created by David Auber, from LaBRI (Laboratoire Bordelais de Recherche en Informatique) and University of Bordeaux, and maintained over the years by an average size development team. Until the 5.1 release, the main Tulip code repository (subversion based) was hosted on SourceForge. That repository remains the Tulip releases download repository but Tulip 5 development are now hosted here.

Tulip is an information visualization framework dedicated to the analysis and visualization of relational data. Tulip aims to provide the developer with a complete library, supporting the design of interactive information visualization applications for relational data that can be tailored to the problems he or she is addressing.

Written in C++ the framework enables the development of algorithms, visual encoding, interaction techniques, data models, and domain-specific visualizations. One of the goal of Tulip is to facilitates the reuse of components and allows the developers to focus on programming their application. This development pipeline makes the framework efficient for research prototyping as well as the development of end-user applications.

The Tulip framework offers numerous features, notably:

• An efficient graph data model in terms of memory usage for storing large networks and the attributes of their elements (called properties in the Tulip semantics). It is also one of the few that offer the possibility to efficiently define and navigate graph hierarchies or cluster trees (nested subgraphs).

• Several graph file formats for serializing such a model to disk, notably the TLP format based on a Lisp syntax for easy parsing but also the TLP binary format for faster graph saving and loading.

• A large variety of graphs algorithms: clustering, metric, layout ... As Tulip is dedicated to graph visualization, it is provided with numerous state of the art graph layout algorithms but also a bridge to the Open Graph Drawing Framework.

• A hardware accelerated graph rendering engine written in OpenGL, highly customizable in terms of visual encoding for graph nodes and edges, in order to efficiently generate aesthetic and interactive visualizations.

• Multiple visualization components (called views in the Tulip semantics) for analyzing graph data using other representations than the classical node-link diagram one: matrix, histograms, scatter plots, parallel coordinates, ...

• Python bindings for the main Tulip C++ API, giving to Tulip scripting facilities for manipulating graphs loaded from its main graphical user interface. The bindings can also be obtained from the Python Packaging Index.

• A plugin based architecture for easily extend the capability of the framework with new graph import mechanisms, graph algorithms, visualization components, ... Tulip plugins can be written in C++ or Python.

• A graphical user interface, based on the Qt framework, enabling to easily interact and manipulate the different components of the framework.

Below are some links to relevant documentation resources about the Tulip framework:

• Tulip User Manual
• Tulip Developer Handbook
• Tulip C++ API
• Tulip Python

Tulip is a cross-platform framework and can be installed or compiled on many Linux distributions (Arch Linux, Debian, Fedora, Gentoo, OpenSUSE, Ubuntu...), on MacOS and Windows.

For each release, Tulip offers precompiled binaries for Linux (using AppImage), MacOS (dmg bundles) or Windows (NSIS based installers). You can download those binaries from SourceForge.

Tulip can be easily compiled on every supported platforms. However, that process can take some times depending on your system configuration.

The following dependencies are required to build Tulip:

• CMake >= 3.1
• A C++11 compiler : GCC >= 4.8.1, Clang >= 3.3 or Microsoft Visual Studio >= 2013
• FreeType
• zlib
• Qt >= 5.6
• OpenGL >= 2.0
• GLEW >= 1.4

In order to build the Python components, the following dependencies are needed:

• Python >= 3.6
• SIP >= 4.19.14 (if SIP can not be found or its version does not meet the minimum requirements, it will be compiled using the existing copy of its source code in the Tulip tree)

The following dependencies are also needed but they will be compiled from the Tulip source tree if they can not be found on your system:

• Qhull
• QuaZIP
• yajl >= 2.0

In order to generate the documentation, the following tools must be installed:

• Sphinx to build the User Manual, Developer Handbook and Python bindings documentation
• Doxygen to build the C++ API documentation

If you are a Linux user, all these dependencies can be installed with the package manager of your distribution. The old continuous integration setup file .travis.yml can give you some hints for Tulip build on Linux platforms.

If you are a MacOS user, we recommend to use MacPorts or Homebrew in order to easily install all these dependencies. Hints on how to build Tulip for these two environments can also be found in the old continuous integration setup file .travis.yml. A dedicated script is also available for MacPorts macports_build.sh.

If you are a Windows user, we recommend to use MSYS2 as it greatly facilitates the build of Tulip on that platform (notably by providing up to date compilers and precompiled dependencies). A dedicated batch file msys2_build.bat is available for MSYS2 build.

Hints on how to build Tulip for MSVC can be found in appveyor_msvc.bat currently used for continuous integration on AppVeyor.

Tulip web site has a dedicated page for Tulip bibliography.

Tulip is free software; you can redistribute it and/or modify it under the terms of the GNU Lesser General Public License as published by the Free Software Foundation, either version 3 of the License, or (at your option) any later version.

Tulip is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License for more details.