Install

Biogeme is an open source software. This page explains how to install it on various platforms. In general, it is recommended to install it from sources, so that the software is tailored to your system. However, for users using Windows, a specific installer is also available.

• Install on Mac OS
• Install an executable on Windows
• Install from sources on Windows
• Install from sources

Biogeme must be installed from the terminal using the package manager homebrew. Make sure to have homebrew installed, following the instructions available here. Once homebrew is installed, type

brew install biogeme

Refer to the homebrew documentation for possible options: click here.

If you do not know how to use a terminal, have a look at the many tutorials on the web, such as this video.

An installer is provided for Windows, that installs an executable for the GUI version. In this case, the additional utilities (such as bioPrepareData) are not available.

Download the installer.

Note that versions of Biogeme running on Windows may be slow. For models requiring a significant computational effort, it is recommended to use Mac OS X or Linux.

The installation from sources (see instructions below) under Windows require the MSYS2 environment. In order to configure it to compile biogeme, follow these instructions.

Install MSYS2

Run the installer x86_64 from msys2.github.io

Open the terminal mingw64 located in directory msys64

Update to the latest version

update-core

You may need to exit the terminal and open a new one

pacman -Sy

You may need again to exit the terminal and open a new one

pacman -Syuu

Install the essentials

pacman --needed -S bash pacman pacman-mirrors msys2-runtime

pacman -Syuu

pacman -S mingw-w64-x86_64-pkg-config

pacman -S base-devel

pacman -S mingw-w64-x86_64-gcc

Install python3

pacman -S mingw-w64-x86_64-python3

Install GTKMM

pacman -S mingw-w64-x86_64-gtkmm3

You can now follow the instructions below to install biogeme from sources.

Note that versions of Biogeme running on Windows may be slow. For models requiring a significant computational effort, it is recommended to use Mac OS X or Linux.

1. Before installing Biogeme from sources, you need to install the following packages on your system:
   • Python 3 (required for PythonBiogeme).
   • GTK MM 3.0 (required for the Graphical User Interface).
   • IPOPT (optional).
2. Download the tarball into a directory of your choice.
3. Unpack the tarball:

   tar xvzf biogeme-2.x.tar.gz

   It will create a directory biogeme-2.x.
4. Connect to the directory biogeme-2.x and type

   ./configure

   By default, it will configure the process to generate BisonBiogeme, PythonBiogeme and the Graphical User Interface, if the proper packages are available on the system. For other options, type

   ./configure --help

   . For instance, it may happen on MSYS2 that biogeme would not link properly with IPOPT. In this case, you can tell biogeme to ignore IPOPT using the following option: ./configure --without-ipopt.
5. Check the output of the configure script to verify that it is configured properly. It should look similar to this:

   biogeme 2.6a for darwin15.6.0: Automatic configuration OK.
Build biogeme............yes
Build biomerge...........yes
Build biosim.............yes
Build python biogeme.....yes (Python version 3.6)
Build GTK GUI version....yes
Use IPOPT optimization...yes
Static executables.......no (use option --enable-staticexec)
Long double..............no (use option --enable-longdouble)
Executables for debug....no (use option --enable-debug)
Versioned installation...no (use option --enable-versioned)


6. Compile the code:

   make -j 8 install

   The option -j 8 speeds up the compilation by using 8 parallel compilation threads.
7. If the default options are selected, the following files are installed:
   • In /usr/local/bin

     biogeme	pybiogeme	biosim
     mod2py	biomerge	weightedhistogram
     histogram	biocheckdata	biopreparedata
     biosampledata	gtkguibiogeme

   • In /usr/local/share

     histogram.py	generateHistogram.py	generateWeightedHistogram.py
     biocheckdata.py	biopreparedata.py	biosampledata.py
     python/bio_expression.py	python/bio_iterator.py	python/biogeme.py
     python/loglikelihood.py	python/statistics.py	python/simulateDiscreteDistribution.py
     python/distributions.py	python/nested.py	python/cnl.py
     python/mev.py	python/boxcox.py	python/legendre.py
     python/bioMatrix.py	python/weightedloglikelihood.py	gtkbiogeme.ui