BioCro-ePhotosynthesis

Coupled BioCro (with the new skeleton module library) and ePhotosynthesis (C++) through linking dynamic libraries.

Tested on MacOS 11.3.1 with x86_64-apple-darwin13.4.0-clang and linux on biocluster with GCC/.8.2.0

On Biocluster, my personal conda environment is called ephoto.

Prerequisites:

• miniconda/anaconda - (Required now since some library paths rely on conda environment variables)
• R
• cmake

Clone this repository

git clone https://github.com/Matthews-Research-Group/BioCro-ePhotosynthesis.git --recursive

If you forgot the --recursive the flag, you can do this to get all submodules,

git submodule update --init --recursive

Building

• First of all, create a conda environment called ephoto Unfortunately, the fixed naming here is required. Usually, I can use $CONDA_PREFIX to get the conda path. However, on clusters, the conda's main path is usually differernt from the user's customized path. Having a fixed environment name makes sure the program can always locate the conda path where we install the specific packages for this project.

• build the ePhotosynthesis C++ version in a conda environment

  #go to models/ePhotosynthesis_C
  mkdir build
  cmake ..
  make
  

  The ePhotosynthesis C++ uses CMAKE to manage the compilation. Please see more details from the README on its Github page to install the package. I suggest to use Conda to do this, which should be much easier. In conda, you can easily install boost, sundials, and cmake, which are required for installing the ePhotosynthesis C++.

  If successful, You should see a file named libePhotosynthesis.dylib or libePhotosynthesis.so in the build folder. This file is needed for the BioCro (ePhotosynthesis) build in the next step.

• build the BioCro and the BioCro (ePhotosynthesis) module library

  We are now using the new skeleton module library framework. This means the BioCro version that couples the ePhotosynthesis is on a separate repository. I personally like it a lot since there are much less files in the folder now.

  To do this, we need to first install the public version of the BioCro. To be consistent with the version that we used in the in silico Plants paper, you may want to check out at c98a8281. You can use git checkout c98a8281 after you cloned the BioCro repository.

   #now go to models/BML-ePhotosynthesis, and run
   R CMD INSTALL .
  

  If successful, you should see two R pakcages named BioCro and BMLePhoto.

Running

Go back to the parent folder BioCro-ePhotosynthesis,

Rscript run_biocro_ephoto.R

This is just an example R script to run scenarios with different climate inputs. You should modify it or have your R script to run your own scenarios.