Version 0.6.5 is packaged up and ready for download on PyPI and conda-forge!

This release improves and expands on the model-wide group constraints added in 0.6.4, further increasing the flexibility they make available.

There is also a new storage_discharge_depth constraint, which allows setting a minimum stored-energy level to be preserved by a storage technology.

This version is also fully Python 3.8 compatible. In the process of updating dependencies for Python 3.8, we updated to the most recent version of scikit-learn (0.22), which fixes a bug in how k-means clustering reacts to specifying a random seed. This may result in models running Calliope 0.6.5 and setting a random seed seeing different k-means clusters than Calliope 0.6.4 and older.

Version 0.6.4 is packaged up and ready for download on PyPI and conda-forge!

Highlights of this version are new group constraints that add a range of flexibility. For example, the cost_var_max group constraint allows flexible emissions limits on a per-region basis (assuming emissions are modelled as a cost class). The demand_share_max or demand_share_per_timestep_max group constraints allow limiting groups of technologies (e.g., variable renewables, or legacy generators) to a maximum on a model-wide basis or for groups of locations, on average or in each individual time step, respectively. Many kinds of policy constraints on system planning can easily be modelled through group constraints.

We have a new logo and website!

Replacing the dusty old static site with little informational value and little visual appeal, the new website is the first landing point for information about Calliope and its applications.

In particular, we now have a model gallery with on existing Calliope models (which are also a good place to start learning when building a new model). More models will be added to the gallery in the near future.