Skip to content

Installation#

There are a few different ways to install Python package for zero- and first-order continuous timeseries. We provide these by use case below. As a short summary, if you:

  • just want to use Python package for zero- and first-order continuous timeseries for validation, and nothing else, go to installation as an application
  • want to use Python package for zero- and first-order continuous timeseries's functionality as a library within another application, go to installation as a library
  • want to develop Python package for zero- and first-order continuous timeseries, go to installation for developers

By use case#

As an application#

If you want to use Python package for zero- and first-order continuous timeseries as an application, then we recommend using the 'locked' version of the package. This version pins the version of all dependencies too, which reduces the chance of installation issues because of breaking updates to dependencies.

The locked version of Python package for zero- and first-order continuous timeseries can be installed with

mamba install -c conda-forge gradient-aware-harmonisation-locked

conda install -c conda-forge gradient-aware-harmonisation-locked

pip install 'gradient-aware-harmonisation[locked]'

As a library#

If you want to use Python package for zero- and first-order continuous timeseries as a library, for example you want to use it as a dependency in another package/application that you're building, then we recommend installing the package with the commands below. This method provides the loosest pins possible of all dependencies. This gives you, the package/application developer, as much freedom as possible to set the versions of different packages. However, the tradeoff with this freedom is that you may install incompatible versions of Python package for zero- and first-order continuous timeseries's dependencies (we cannot test all combinations of dependencies, particularly ones which haven't been released yet!). Hence, you may run into installation issues. If you believe these are because of a problem in Python package for zero- and first-order continuous timeseries, please raise an issue.

The (non-locked) version of Python package for zero- and first-order continuous timeseries can be installed with

mamba install -c conda-forge gradient-aware-harmonisation

conda install -c conda-forge gradient-aware-harmonisation

pip install gradient-aware-harmonisation

Additional dependencies can be installed using

If you are installing with mamba, we recommend installing the extras by hand because there is no stable solution yet (see conda issue #7502)

If you are installing with conda, we recommend installing the extras by hand because there is no stable solution yet (see conda issue #7502)

# To add plotting dependencies
pip install 'gradient-aware-harmonisation[plots]'

# To add scipy (an optional dependency)
pip install 'gradient-aware-harmonisation[scipy]'

# To add all optional dependencies
pip install 'gradient-aware-harmonisation[full]'

For developers#

For development, we rely on uv for all our dependency management. To get started, you will need to make sure that uv is installed (instructions here (we found that the self-managed install was best, particularly for upgrading uv later).

For all of our work, we use our Makefile. You can read the instructions out and run the commands by hand if you wish, but we generally discourage this because it can be error prone. In order to create your environment, run make virtual-environment.

If there are any issues, the messages from the Makefile should guide you through. If not, please raise an issue in the issue tracker.

For the rest of our developer docs, please see development.