Installation

This package is available through PyPI, so it can be installed through pip:

pip install heightmap-interpolation

Installation from sources

This project expects python3.7 (or above).

Start by cloning the project:

git clone https://github.com/coronis-computing/heightmap_interpolation.git

We provide the requirements of this project in a “requirements.txt”

If you want to do it in a virtual environment:

cd <path_to_this_project>
python -m venv venv
source venv/bin/activate
pip install -r requirements

To ease the calls to the main interpolate_netcdf4.py script, you can add the container folder to the PATH:

export PATH=$PATH:<path_to_this_package>/heightmap_interpolation/apps

In Ubuntu (and other linux distros), you can set this command as a new line in your <home>/.bash.rc for these changes to persist on new terminals.

Alternative: use the pre-compiled docker

For convenience, we also provide a docker image with all the dependencies installed at DockerHub. Assuming you have docker installed, you can obtain it by:

docker pull coroniscomputing/heightmap_interpolation:<tag_name>

Where <tag_name> must be a specific version of the package, or latest.

Then, run it with (tested in Ubuntu):

docker run -it --user $(id -u):$(id -g) -v <data_folder>:/data coroniscomputing/heightmap_interpolation:<tag_name>

On the one hand, using the -v flag we are mounting the directory containing the data to process to the /data folder within the container. On the other hand, the --user $(id -u):$(id -g) part is to achieve that the files you generate within docker in the mounted volume are owned by your user (otherwise they would be owned by the root user!).

The container will automatically run the bash command, and you will be inside the container. Thus, there you can simply run the interpolate_netcdf4.py script with the desired parameters (it is included in the container’s path, so you can call it directly). For instance:

interpolate_netcdf4.py -o /data/<netcdf_results_file> linear /data/<netcdf_input_file>

Keep in mind that this way of running the docker does not provide visualization, so the --show flag will be useless! There are ways of sharing the Xs with docker, but these are out of the scope of this documentation.