Installation

Prerequisites for basic installation

To install GASTLI, you need to:

• have a Python 3.9+ installation,

• have a fortran compiler, for example gfortran,

• download the zip file for the input data here, and place it in a directory in your computer that is easy to find. See the subsection “Setting the input data path” below for instructions on how to specify the path of this directory.

If you do not have Fortran installed, follow the Fortran installation instructions appropriate for your OS. We offer an installation guide for different OS below. Once Fortran is installed, you can then proceed with the GASTLI package installation.

Fortran installation

Warning

Mac OS with Apple Silicon processors (M1, M2 and M3) have ARM64 as the native architecture. These systems have shown problems when installing a Python package that compiles a fortran backend because Python’s default fortran compiler uses Intel architecture. One way to recognise this is to try installing GASTLI with pip (or from source) and testing the installation as indicated below. Users that get a message such as the following need to follow the instructions for Apple Silicon in the Fortran installation in Mac OS (Apple Silicon M1/M2/M3) section:

Traceback (most recent call last):
File "<stdin>", line 1, in <module>
ImportError: dlopen(/Users/acuna/anaconda3/envs/GASTLI_norosetta/lib/python3.10/site-packages/gastli/dimensions.cpython-310-darwin.so, 0x0002): symbol not found in flat namespace (_f2pyinitdimensions_)

• Fortran installation in Linux
• Fortran installation in Mac OS (Intel)
• Fortran installation in Mac OS (Apple Silicon M1/M2/M3)
  • 1) Installation with Rosetta (recommended)
  • 2) Installation with disk image

Installation of GASTLI’s Python package

Our recommendation is to use a Python virtual environment such as venv or conda, to prevent potential conflicts.

If you decide to create a new conda environment (for example, “GASTLI_env”), run in the terminal:

conda create -n GASTLI_env python=3.10

Then activate the environment:

conda activate GASTLI_env

If instead you want to create a virtual environment, use the following command. This will create a new directory that contains all the files necessary to run the environment:

python3 -m venv GASTLI_env

Then activate it with:

source GASTLI_env/bin/activate

Once the environment is activated, we will be ready to install GASTLI’s Python package. This can be done with pip:

pip install gastli

Or for a particular version of the package:

pip install gastli==0.9.3

The package can also be installed from source. This can be done by downloading the .zip file from the GASTLI github repository while making sure you download the intended branch - main, or development - or using git to clone the repository in your local computer as explained here:

git clone https://github.com/lorenaacuna/GASTLI

To move to the root directory and install the package, run in your terminal:

cd GASTLI
pip install .

Setting the input data path

After downloading the input data file (see “Prerequisites for basic installation”), place the gastli_input_data folder in any directory on your computer - just make sure you remember where, such as your Desktop.

Next, set the environment variable GASTLI_input_data_path to your .bash_profile, .bashrc, or .zshrc file, depending on your operating system and shell type. This variable should contain the path to the gastli_input_data folder on your computer. Here are examples of how to do this:

• On macOS: Run the following command in your terminal:

echo 'export GASTLI_input_data_path="/PATH/TO/YOUR/FOLDER/gastli_input_data/"' >>~/.bash_profile

• On Linux: Run the following command in your terminal:

echo 'export GASTLI_input_data_path="/PATH/TO/YOUR/FOLDER/gastli_input_data/"' >>~/.bashrc

• On a SLURM system: Add the following line to your .sh script:

export GASTLI_input_data_path="/PATH/TO/YOUR/FOLDER/gastli_input_data/"

You only need to add this line to your bash profile once.

Alternatively, you can define the environment variable directly within your Python scripts. If you choose this method, add the following line to every script where you use GASTLI’s functions or classes:

os.environ['GASTLI_input_data_path']="/PATH/TO/YOUR/FOLDER/gastli_input_data/"

Testing the installation

You can check quickly if GASTLI installed correctly by running this in Python:

import gastli.dimensions as dim
print(dim.dimensions.n_lay)
import gastli.constants as cte
print(cte.constants.f_alloy_e)

The output should be:

3
0.13