For quite a while now we have been working on a major new version of Calliope - v0.7.0 – packed with powerful new features and major usability upgrades.

A pre-release version of v0.7.0 is already available through both PyPI and Conda-Forge. For new projects, we recommend using this new version already as it introduces a range of backwards-incompatible changes. The Calliope team is using it productively in several projects.

Some of the major new features include:

• Easy customisation of all model math: Extending our philosophy of easy customisation further, Calliope now allows full customisation of the model math. We have built a math parser and all of the pre-defined math inside Calliope is formulated in our own YAML-based math notation, which is easy to read, understand and modify. The documentation on defining your own math explains how you can make use of this powerful functionality to build your own custom constraints, or modify the build-in ones.
• Ability to load any data from tabular files, not just time series.
• Direct interface to the Gurobi API: This speeds up model generation and requires less memory.
• Improvements and streamlining for the terminology and to component definitions (for example, all technologies can now include a storage buffer).
• Visualisation has been removed from Calliope itself; instead we have a dedicated interactive visualisation tool called Calligraph which is also available as a pre-release. This is a preview of Calligraph in action:

Much of the internal code has been re-built around these new features. In this process, we have also made a large number of smaller changes, for example, an easier way to access shadow prices.

See the (also completely re-built) documentation for v0.7.0 by going to the “latest” branch on ReadTheDocs.

The docs contain detailed guidance on migrating a model from v0.6 to v0.7, which also gives you a fuller picture of the many improvements in the new version.

We welcome feedback and contributions!