Skip to content
The ix modeling platform for integrated and cross-cutting scenario analysis
Branch: master
Clone or download
Type Name Latest commit message Commit time
Failed to load latest commit information.

The ix modeling platform (ixmp)


The ix modeling platform (ixmp) is a data warehouse for high-powered scenario analysis, with interfaces to Python and R for efficient scientific workflows and effective data pre- and post-processing, and a structured database backend for version-controlled data management.

This repository contains the core and application programming interfaces (API) for the ix modeling platform (ixmp), as well as a number of tutorials and examples for a generic model instance based on Dantzig's transport problem.


Copyright 2017-18 IIASA Energy Program

The platform package is licensed under the Apache License, Version 2.0 (the "License"); you may not use the files in this repository except in compliance with the License. You may obtain a copy of the License at

Please refer to the NOTICE for details and the user guidelines.

Documentation and tutorial

Documentation of ixmp and the MESSAGEix framework is automatically created from the documentation of the Python and R API packages. The online documentation is synchronyzed with the contents of the master branch of the repositories and

There are a number of tutorials to get started with ixmp. You may want to try the tutorial/transport...

Follow the instructions in doc/README for building the ixmp documentation including the scientific programming API manuals on your local machine.

Scientific reference

Please cite the following manuscript when using the MESSAGEix framework and/or the ix modeling platform for scientific publications or technical reports:

Daniel Huppmann, Matthew Gidden, Oliver Fricko, Peter Kolp, Clara Orthofer, Michael Pimmer, Nikolay Kushin, Adriano Vinca, Alessio Mastrucci, Keywan Riahi, and Volker Krey.
"The |MESSAGEix| Integrated Assessment Model and the ix modeling platform".
Environmental Modelling & Software 112:143-156, 2019.
doi: 10.1016/j.envsoft.2018.11.012
electronic pre-print available at

Install from Conda (New Users)

  1. Install Python via Anaconda. We recommend the latest version, e.g., Python 3.6+.

  2. Install GAMS. Importantly:

    • Check the box labeled Use advanced installation mode
    • Check the box labeled Add GAMS directory to PATH environment variable on the Advanced Options page.
  3. Open a command prompt and type

    conda install -c conda-forge ixmp

Install from Source (Advanced Users)


  1. Install the Java Development Kit (Java SE 8) and set the environment variable JAVA_HOME per the JDK website instructions; if JAVA_HOME does not exist, add as new system variable. At this point, ixmp is not compatible with JAVA SE 9.

  2. Update your PATH environment variable to point to the JRE binaries and server installation (e.g., C:\Program Files\Java\jdk[YOUR JDK VERSION]\jre\bin\, C:\Program Files\Java\jdk[YOUR JDK VERSION]\jre\bin\server).

    Do not overwrite the existing `PATH` environment variable, but add to the list of existing paths.
  3. Install the latest version of GAMS, otherwise, the export to GDX may not work properly (visit If you only have a license for an older verson of GAMS, install both versions. Note that for using the integrated MESSAGEix-MACRO model, it is important to install a GAMS version >= 24.8.1.

  4. Update your PATH environment variable to point to the GAMS installation (e.g. C:\GAMS\win64\24.8); again do not overwrite existing PATH but rather add to end.

  5. Install some version of Python (2.7 is supported, but 3.6 or higher is recommended). Anaconda is a good choice for users not yet familiar with the language (during installation select add anaconda to PATH system variable)

  6. Install a Windows C++ Compiler

  7. Install a version of git, (see, e.g., the website)

Additional dependencies for R users

  1. Make sure the R version installed is either 32 OR 64 bit (and >= 3.3.0), consistently with GAMS and Java. Having both 32 and 64 bit generates error.

  2. Install packages, devtools and optparse via the R package manager

  3. Two alternative options are available:

    • use the package developed in Python, requires the reticulate R package (it will allow to install/use retixmp)

    • use specific R packages (less available functionalities), requires the rJava R package (it will allow to install/use rixmp)

  4. Install Rtools and add the path to the environment variables

  5. For working with Jupyter notebooks using R, install the IRkernel.

Installing the ix modeling platform

  1. Fork this repository and clone the forked repository (<user>/ixmp) to your machine. To fork the repository, look for the fork button in the top right at iiasa/ixmp. Add iiasa/ixmp as upstream to your clone.

    We recommend GitKraken for users who prefer a graphical user interface application to work with Github (as opposed to the command line).

  2. Open a command prompt in the new ixmp directory and type:

    $ pip install .
  3. (Optional) Run tests to check that ixmp works on your system:

    $ pip install .[tests]
    $ py.test tests

Notes and Warnings

  1. For Anaconda users experiencing problems during installation of ixmp, Anaconda might not have been added to the PATH system variable properly. So, if install.bat just opens and collapses again, check if:

    C:\[YOUR ANACONDA LOCATION]\Anaconda3\Scripts;
    C:\[YOUR ANACONDA LOCATION]\Anaconda3\Library\bin;

    are all part of the PATH system variable. If they are not there, add them.

Building Documentation

  1. Install dependencies for the documentation:

    $ pip install .[doc]
  2. Navigate to the doc directory and build the documentation:

    $ cd doc
    $ make html

The built documentation appears in doc/build/html/index.html; open this file using a web browser.

You can’t perform that action at this time.
You signed in with another tab or window. Reload to refresh your session. You signed out in another tab or window. Reload to refresh your session.