Link Search Menu Expand Document

Conda installation

The best and the easiest way to install ASE/GPAW/Elastic system with gpaw-tools is a conda installation. However, sometimes GPAW can run up to 20 times slower with conda installation. Please keep this in mind. Download and install the miniconda. You can say ‘yes’ or ‘no’ to initialization after installing it:

$ wget https://repo.anaconda.com/miniconda/Miniconda3-latest-Linux-x86_64.sh
$ chmod +x Miniconda3-latest-Linux-x86_64.sh
$ ./Miniconda3-latest-Linux-x86_64.sh

then you can update miniconda:

$ eval "$(/home/$USER/miniconda3/bin/conda shell.bash hook)"
$ conda update conda

Now, we can create an environment (here ‘gpaw-env’ name is used. You can use any name) and activate it:

$ conda create --name gpaw-env
$ conda activate gpaw-env

Then, install GPAW and Elastic packages

$ conda install -c conda-forge gpaw elastic requests phonopy pandas

NOTE: For the energy consumption measurement feature, you can install mongo and pandas packages with conda. However, you must install only pyRAPL with pip in the next step. This workflow and the energy consumption measurement feature is not tested on conda yet. Please submit a better installation solution in this page. Thanks.

NOTE: If the user wants to use gpaw-tools versions 23.7.0 and before, GPAW version 22.8.0 must be used. The above command will be conda install -c conda-forge gpaw=22.8.0 elastic requests

Lastly, download and install gpaw-tools.

NOTE: Sometimes, after this step, mpirun -np <core_number> command.py runs command.py seperately in a number of core_number instances. If you observe this kind of execution, please deactivate conda deactivate then remove environment conda remove --name gpaw-env --all then create same environment again conda create --name gpaw-env then activate it conda activate gpaw-env and then install all files again conda install -c conda-forge gpaw elastic requests. Problem must be solved. If it is not solved, please open a new issue about your problem.