NED-toolbox

project code for GreenHEART scaled to the national scale

• Install on Local Machine
• Install on HPC
• Quick Tour

Install on Local Machine

1. navigate to local target directory and clone the repository, then navigate to NED-toolbox

   cd /path/to/target/dir
   git clone https://github.com/elenya-grant/NED-toolbox.git
   cd NED-toolbox
   

2. Create and activate conda environment (named ned_tools)

   conda create --name ned_tools python=3.8 -y
   conda activate ned_tools
   

3. Install dependencies

   conda install -c conda-forge mpi4py petsc4py
   pip install -r requirements.txt
   

4. (option 1) install greenheart using pip

   pip install -r requirements-dev.txt
   

5. (option 2) install greenheart using git

   git clone -b feature/ned https://github.com/elenya-grant/HOPP.git
   cd HOPP
   conda install -c conda-forge coin-or-cbc=2.10.8 -y
   conda install -c conda-forge glpk -y
   pip install -r requirements.txt
   pip install -e .
   cd
   cd /path/to/target/dir/NED-toolbox
   

6. Install NED-toolbox

   pip install -e .
   

7. Set environment for aggregated results folder used for preprocessing optimization runs. In the root directory, make a file titled ".env" and add the line MAIN_RESULTS_FOLDER=/path/to/where/results/are.

Optional: install GreenHEART more locally

do something

Install on HPC

• setup github ssh on the HPC to use these instructions

Install NED-toolbox

1. navigate to HPC target directory (such as /scratch/) and clone the repository (or your fork of NED-toolbox), then navigate to NED-toolbox

cd ../../scratch/<user>
git clone [email protected]:elenya-grant/NED-toolbox.git
cd NED-toolbox

2. Create and activate conda environment

module load conda
export CONDA_PKGS_DIRS=/scratch/<user>/.conda-pkgs
conda create --prefix /scratch/<user>/ned_tools python=3.8
conda activate /scratch/<user>/ned_tools

3. install mpi4py

   module load cray-mpich
   

   • check the mpicc, following command should return /opt/cray/pe/mpich/8.1.28/ofi/crayclang/16.0/bin/mpicc

     which mpicc
     

   • install mpi4py

     python -m pip install mpi4py
     

4. install dependencies

pip install -r requirements.txt

5. install GreenHEART (see options in local install)
6. finalize setup

pip install -e .

Quick Tour