Downloading libSBML

Please use the following table to find the appropriate instructions for your situation:

1. MATLAB 2. Python 2.1 Using Conda 2.2 Using PyPI 2.3 Using Linux's apt-get and yum together with online repositories 2.4 Using the Windows libSBML installer for Python 3. C, C++, C#, Java, JavaScript, Perl, PHP or Ruby 4. R 4.1 Using the R GUI 4.2 Using a shell command 5. Other languages

1. MATLAB

For stable releases of libSBML, we provide a unified installer for all versions of MATLAB and all operating systems. The result is ready to run after installation; it does not need compilation, and only requires you tell MATLAB where you placed the files. Here are the steps:

1. Download the distribution archive from SourceForge.
2. Extract the archive somewhere on your local computer.
3. Start your copy of MATLAB, then either:
   1. Add that folder to your MATLAB path using MATLAB's addPath command.
   2. Verify it works: change into that folder within your running MATLAB environment and issue the command TranslateSBML('test.xml'). If all is well, the output will be a MATLAB data structure.
   3. Save the changes to your MATLAB path for future use. You can use MATLAB's standard savepath command.
   or
   1. Navigate to the directory where you extracted the archive.
   2. Run installSBML, which performs the above steps.

Note that the MATLAB installation archive contains only the MATLAB interface; it does not contain the core libSBML library or language libSBML API interfaces for other programming languages.

2. Python

We provide four ways to obtain the libSBML Python interface. Please use whichever one you prefer.

2.1 Using Conda

We provide a distribution of the libSBML Python interface through Anaconda Cloud, making it accessible through conda. To install it this way, follow these instructions:

• To get the stable version of libSBML:

  conda install -c SBMLTeam python-libsbml
  

• To get the experimental version of libSBML:

  conda install -c SBMLTeam python-libsbml-experimental
  

After this, you should be able to access the Python language interface for libSBML directly from your Python interpreter by importing it like any other Python module. In other words, import libsbml should work in your Python interpreter.

2.2 Using PyPI

Important: the default version of pip on your system may not be appropriate for the version of Python you want to use. (For example, the default pip may be for Python 2.7, which will not work if you want to use Python 3.x.) If you do not have pip installed for a given version of Python, a reliable way to install it is the following:

1. Download get-pip.py.
2. Run the following command in a shell terminal, where version is your version of Python (e.g., 3):

   sudo pythonversion get-pip.py
   

   An example of the command line above is the following: sudo python3 get-pip.py.

Assuming you have either pip2 or pip3 on your system, then to get the stable release of libSBML, execute the following command:

sudo pipversion install python-libsbml

where version is either 2 or 3 depending on whether you want to use Python 2 or Python 3. Alternatively, for the experimental release of libSBML, use the following command instead:

sudo pipversion install python-libsbml-experimental

Note that in all cases, the process will take a long time because it must install and build dependent libraries. If all goes well, the steps above should result in pip eventually printing the following message to tell you has it finished successfully:

[... much text omitted ...]
Successfully installed python-libsbml
Cleaning up...

After this, you should be able to access the Python language interface for libSBML on your computer directly from your Python interpreter by importing it like any other Python module. In other words, import libsbml should work when typed into your Python interpreter.

2.3 Using Linux's apt-get or yum

If you are running Linux, and you prefer to use the standard over-the-air installation tools provided by your operating system (e.g., yum and apt-get), you can take advantage of binary packages we provide using the OpenSUSE Build System (OBS) for many different flavors of Linux including CentOS, Fedora, Ubuntu, and others:

• For the stable release: python-libsbml.
• For the experimental release: python-libsbml-experimental.

As an example, the following is a screenshot of the page currently found at the python-libsbml page. From this page, clicking on the name of an operating system will display the instructions for what to do:

Screenshot of the OBS python-libsbml page for different Linux systems.

After following the instructions to install the appropriate binary distribution on your computer, you should be able to access the Python language interface for libSBML on your computer directly from your Python interpreter. In other words, import libsbml should work when typed into your Python interpreter.

2.4 Using the Windows libSBML installer for Python

You can take advantage of executable installers that we make available for that platform. They can be found in a subdirectory named python within each of the Windows subdirectories (stable and experimental) on SourceForge. There are different installers for different versions of Python. The following screenshot shows an example.

Screenshot of the SourceForge download page for the 64-bit Windows stable libSBML installers.

After you download an appropriate .exe file for your system, run the installer. The installer will lead you through the installation process.

3. C, C++, C#, Java, JavaScript, Perl, PHP or Ruby

The installation packages in the libSBML SourceForge download site contain interfaces for C, C++, C#, Java, JavaScript, Perl, PhP, Python, and Ruby in the same package. (However, there are better ways to get the Python interface, as explained above.) The download area has two subdirectories: one for stable releases of libSBML, and one for experimental releases. The stable releases contain support for only the accepted parts of SBML, while the experimental releases contain code for SBML Level 3 packages still in development.

Screenshot of the SourceForge download page for libSBML.

Click on either stable or experimental, and then navigate to a subdirectory matching your operating system platform. Download the appropriate installer or package file for your platform, and use the appropriate platform-specific approach to install the software. (E.g., if you downloaded a Mac .dmg file, double-click the file in the Mac Finder to mount the disk image, open the image contents in the Finder, and finally, double-click the .pkg file contained inside.)

4. R

We provide installers for the stable version of the libSBML R language interface for Linux, Mac OS X and Windows. Download the version that is appropriate for your situation:

• For Linux and Mac OS X: Download the libSBML R source package from SourceForge.
• For Windows: Download the 64-bit version or the 32-bit version of the binary installer for R.

Once you have downloaded the appropriate version, one more step is required to install the libSBML R interface. You can do this step in one of two ways: either using the R graphical interface, or using a terminal/shell command. These approaches are explained below.

4.1 Using the R GUI

From the R menu "Packages" or "Packages & Data" (depending on your operating system), select the "Package Installer" item. You will be presented with an interface that lets you install a local binary package. Use that interface to navigate to the directory where you copied the libSBML archive file, and select the archive file for installation.

4.2 Using a shell command

Suppose that the path to the libSBML R archive on your computer is archive. Execute the following command in a terminal/shell window (note: not in the R command line interpreter!):

R CMD INSTALL archive

If all goes well, you should see R print messages. If you're using the source package, it will be something such as the following (the paths on your computer may differ):

* installing to library '/Library/Frameworks/R.framework/Versions/3.1/Resources/library'
* installing *binary* package 'libSBML' ...
* DONE (libSBML) 

whereas if you are using a binary package, the messages will be something like the following instead:

> utils:::menuInstallLocal()
  package ‘libSBML’ successfully unpacked and MD5 sums checked
>

From this point forward, you should be able to call on libSBML functions from within your copy of R. You can test whether the installation was successful by trying the following command in R:

library('libSBML')

If R does not report an error when you execute this command, then the libSBML R interface is probably installed correctly.

5. Other languages

If no installation package is available for your particular platform, or you have special requirements for libSBML's configuration, you can always build libSBML from sources. The process for doing that is the topic a separate section, Building and installing libSBML .