3.0.1-dev0
 EDK External drift kriging.
Compilation

The section 'Dependencies' lists the general requirements for the compilation. The section 'System-dependent dependency installation' gives some instructions on how to install these dependencies on Windows, some Linux-distributions and MacOS. Conda based dependency installation is described in the section 'Conda dependent installation'.

# Dependencies

To clone and compile EDK you need at least the following:

# System-dependent dependency installation

After you installed all dependencies on your system you can proceed with cloning and compiling.

## Unix (Linux / MacOS)

1. MacOS with homebrew available:
brew install git gcc netcdf lapack cmake
2. Ubuntu, Mint and other apt-get based systems with matching repositories:
sudo apt-get install git gfortran netcdf-bin libnetcdf-dev libnetcdff-dev liblapack-dev cmake
3. Archlinux:
sudo pacman -S git gcc-libs netcdf-fortran lapack cmake
4. yum based systems (CentOS, OpenSuse):
sudo yum -y install git gcc-gfortran netcdf-fortran lapack cmake

## Windows

On Windows 10 and later we recommend to use Windows Subsystem for Linux (WSL) to be able to use Linux by e.g. installing Ubuntu there.

Easiest way to do so is:

1. install the Windows Terminal
2. open the Windows Terminal and type:
wsl --install -d ubuntu
3. Open Ubuntu from the new entry in the start menu

Then you can follow the install instructions for Ubuntu from above.

## Module systems

If you are on a module system, load the modules gcc or intel depending on your favorite compiler. Then, load the modules netcdf-fortran and cmake.

These modules will have system specific names, environments, etc. You may use module spider to find the right packages and the right dependencies, potentially use corresponding wiki pages.

### On EVE (the cluster at the UFZ)

A set of load-scripts is provided in hpc-module-loads (see the repository for more details), to load all needed modules for specific compilers:

• Example: GNU 7.3 compiler (foss/2018b Toolchain):
or (MPI support)

# Conda dependent installation

The simplest way to compile this project on your local computer is to use a conda environment (on Linux (including WSL) or MacOS) provided by Miniforge to install NetCDF, a Fortran compiler, make, cmake.

On Windows we recommend to use Windows Subsystem for Linux (WSL) to be able to use Linux (see above) and set up conda there.

wget https://github.com/conda-forge/miniforge/releases/latest/download/Miniforge3-$(uname)-$(uname -m).sh
bash Miniforge3-$(uname)-$(uname -m).sh

To create a (local) conda environment with all dependencies type the following:

conda create -y --prefix ./fortran_env
conda activate ./fortran_env
conda install -y git cmake make fortran-compiler netcdf-fortran liblapack

Then you can proceed with cloning and compiling.

# Cloning the repository

First you need to clone the repository (if you already have git, otherwise see below):

git clone https://git.ufz.de/chs/progs/edk_nc.git

This will give you a new folder edk_nc/ containing the whole repository. You can go into it by:

cd edk_nc/

If you then want to compile a specific version (different from the latest development version), you can check that out with e.g.:

git checkout v3.0.0

Afterwards you can continue with the compilation.

# Compilation commands

It could be necessary to set your desired fortran compiler with an environment variable, e.g.:

export FC=gfortran

We prepared a set of scripts to automatize the build and compilation process to generate an executable in the root directory with the following naming scheme:

• Release version edk:
source scripts/compile
• Debug version edk_debug:
source scripts/compile_debug
• Release version with OpenMP support edk_openmp:
source scripts/compile_OpenMP
• Debug version with OpenMP support edk_openmp_debug:
source scripts/compile_OpenMP_debug

Then you can find an executable edk (or edk[_openmp][_debug]) in the current folder. You can execute it with:

./edk

# Installation

To install edk after compilation, i.e. make it available as a command edk, you can do the following (assuming you used the release compile script, otherwise replace release with the respective build folder):

cmake --install release

If you need to provide a prefix, where to install it, you can just pass one. For example, if you used a conda environment for compilation, you can also install edk there with:

cmake --install release --prefix \$CONDA_PREFIX

# Compilation without Internet

Starting with version 3.0, EDK is depending on FORCES, our Fortran library for Computational Environmental Systems. This library is downloaded on the fly by CPM, the cmake package manager.

If you don't want to download it indirectly, know you wont have internet during your development or you want to work on routines provided by FORCES, you can place a copy of the FORCES repository in the root of your cloned EDK repository by e.g.:

git clone https://git.ufz.de/chs/forces.git

The new folder forces/ will be automatically recognized during compilation as described above and nothing will be downloaded.

If you just want a specific version (see src/CMakeLists.txt for the currently used one), do this:

git clone --branch v0.3.2 --depth 1 https://git.ufz.de/chs/forces.git

If you have already cloned FORCES somewhere else, you can also provide a path to this repository. You can do this with all mentioned compile scripts, e.g.:

source scripts/compile -DCPM_forces_SOURCE=<path/to/your/forces/repo>

For example, if you have cloned FORCES next to edk_nc, this could look like this:

source scripts/compile -DCPM_forces_SOURCE=../forces

The presented compile scripts all just execute two cmake commands with a specific set of configuration flags. The basic cmake workflow, to configure and compile in a build/ folder, is:

cmake -B build
cmake --build build

You can control all cmake options by passing them as directives staring with -D to the cmake configuration. For example for debug configuration, you can do the following:

cmake -B build -DCMAKE_BUILD_TYPE=Debug

To configure the build interactively, you can also use ccmake (command line tool) or the CMake GUI (graphical user interface). Check their respective documentation for further information. To use ccmake you can do the following:

cmake -B build
ccmake build

Then set your desired options and re-configure your build (by pressing c). Afterwards build you project as always by executing:

cmake --build build --parallel