Installation Instructions

DYCOVE is compatible with Python 3.10-3.13 (see Issues or notes below for some current incompatibilities). There are three general dependencies, which are numpy, xarray, and scipy. However, there are additional dependencies depending on the underlying numerical model to be used, and if the ModelPlotter will be used.

Install ANUGA

For DYCOVE with ANUGA, anuga [1] must be installed. ANUGA may have additional dependencies depending on the application, such as mpi4py for running in parallel. The easiest way to install ANUGA is to create a new conda environment and install via conda-forge (from the ANUGA documentation page):

conda install -c conda-forge anuga mpi4py

Update: ANUGA is not currently importing with Python 3.12 and is causing Python to crash without an error message. For now, use Python 3.10, 3.11, or 3.13, and let us know if it starts working again with 3.12 by responding to the relevant Issue on GitHub.

Install Delft3D FM

For DYCOVE with Delft3D FM, the software itself [2] will of course need to be licensed and installed locally. From a Python standpoint, it is also recommended to create a new environment to install DYCOVE and its dependency bmi-python, which provides a Basic Model Interface [3] between Python and the Delft3D FM software. From a dedicated environment, the easiest way to install bmi-python is from source:

git clone https://github.com/openearth/bmi-python.git
cd bmi-python
pip install -e .

Windows users may also need to install pypiwin32:

pip install pypiwin32

Update: The DYCOVE implementation of the DFM bmi library is not currently working with Python 3.11 and 3.12.

Install Plotting Libraries

DYCOVE’s ModelPlotter is a convenient way to plot outputs. The additional dependencies are matplotlib, tqdm, and imageio. Of course, users will want to have some of these libraries installed whether or not they plan to use ModelPlotter.

Installing DYCOVE

DYCOVE itself can be installed from source like any other package:

git clone https://github.com/Ecomomo-lab/dycove-model.git
cd dycove-model
pip install -e .

References