jarvis-tools 2019.7.31

High throughput computation with density functional theory, molecular dynamics and machine learning. https://jarvis.nist.gov/

JARVIS

Joint Automated Repository for Various Integrated Simulations (JARVIS) is an integrated framework for computational science using density functional theory, classical force-field/molecular dynamics and machine-learning. The jarvis-tools package can be used for high-throughput computation, data-analysis, and training machine-learning models. Some of the packages used in the jarvis-tools package are shown below. JARVIS-official website: https://jarvis.nist.gov

Installing JARVIS

• We recommend installing miniconda environment from https://conda.io/miniconda.html

  bash Miniconda3-latest-Linux-x86_64.sh (for linux)
  bash Miniconda3-latest-MacOSX-x86_64.sh (for Mac)
  Download 32/64 bit python 3.6 miniconda exe and install (for windows)
  Now, let's make a conda environment just for JARVIS::
  conda create --name my_jarvis python=3.6
  source activate my_jarvis
  

• Git clone install (Recommended):

  pip install numpy scipy matplotlib
  git clone https://github.com/usnistgov/jarvis.git
  cd jarvis
  python setup.py install
  

• Alternative pip install:

  pip install numpy scipy matplotlib
  pip install jarvis-tools
  

• Alternative nix install:: Nix allows a robust and reproducible package for Linux. To generate a Nix environment for using JARVIS, follow the Nix instructions.

Jupyter notebooks

• Python for beginners:

• JARVIS-DFT data analysis:

• JARVIS-ML training:

• Comparing ML algorithms:

• JARVIS-FF data-analysis:

• See more in the plot-gallery below

References

• JARVIS-FF:

  1. Evaluation and comparison of classical interatomic potentials through a user-friendly interactive web-interface, Nature: Sci Data. 4, 160125 (2017).https://www.nature.com/articles/sdata2016125
  2. High-throughput assessment of vacancy formation and surface energies of materials using classical force-fields, J. Phys. Cond. Matt. 30, 395901(2018).http://iopscience.iop.org/article/10.1088/1361-648X/aadaff/meta

• JARVIS-DFT:

  1. High-throughput Identification and Characterization of Two-dimensional Materials using Density functional theory, Scientific Reports 7, 5179 (2017).https://www.nature.com/articles/s41598-017-05402-0
  2. Computational Screening of High-performance Optoelectronic Materials using OptB88vdW and TBmBJ Formalisms, Scientific Data 5, 180082 (2018).https://www.nature.com/articles/sdata201882
  3. Elastic properties of bulk and low-dimensional materials using van der Waals density functional, Phys. Rev. B, 98, 014107 (2018).https://journals.aps.org/prb/abstract/10.1103/PhysRevB.98.014107
  4. Convergence and machine learning predictions of Monkhorst-Pack k-points and plane-wave cut-off in high-throughput DFT calculations, Comp. Mat. Sci. 161, 300 (2019).https://www.sciencedirect.com/science/article/pii/S0927025619300813?via%3Dihub
  5. High-throughput Discovery of Topologically Non-trivial Materials using Spin-orbit Spillage, Nature: Sci. Rep. 9, 8534,(2019), https://www.nature.com/articles/s41598-019-45028-y
  6. Accelerated Discovery of Efficient Solar-cell Materials using Quantum and Machine-learning Methods, Chem. Mater., https://pubs.acs.org/doi/10.1021/acs.chemmater.9b02166
  7. Data-driven Discovery of 3D and 2D Thermoelectric Materials , https://arxiv.org/abs/1903.06651.

• JARVIS-ML:

  1. Machine learning with force-field inspired descriptors for materials: fast screening and mapping energy landscape, Phys. Rev. Mat., 2, 083801 (2018).,https://journals.aps.org/prmaterials/abstract/10.1103/PhysRevMaterials.2.083801
  2. Materials science in the artificial intelligence age: high-throughput library generation, machine learning, and a pathway from correlations to the underpinning physics, MRS Comm., 1-18 https://doi.org/10.1557/mrc.2019.95

Pypi, Readthedocs and Slideshare links

https://pypi.org/project/jarvis-tools

https://jarvis-tools.readthedocs.io/en/latest/

https://www.slideshare.net/KAMALCHOUDHARY4

Running the examples

• For running high-throughput calculations, set HPC/system related information in env_variables
• Run py.test in tests folder to ensure basic setup

• LAMMPS example:

  An example calculation for Aluminum is given in the lammps folder for running EAM calculation (https://github.com/usnistgov/jarvis/blob/master/jarvis/lammps/examples/basic_input_output.py). Untar the example folder using tar -xvzf Al03.eam.alloy_nist.tgz . Change the ‘parameters’ variable and run jlammps.py.

• VASP example:

  Similarly, an example calculation for Silicon is given in vasp folder (https://github.com/usnistgov/jarvis/blob/master/jarvis/vasp/examples/runstruct_pyvasp.py). The input is a POSCAR file, which is already provided. executable paths, pseudopotential directory path and Special_POTCAR.yaml path needs to be adjusted in joptb88vdw.py top section. The master.py can be submitted to the queuing system with qsub sub.sh.

• ML example:

  We trained machine learning models using JARVIS-DFT data on bandgaps, formation energies and elastic modulus and other properties. We used both chemical and structural descriptors during GradientBoostingRegression training. Example of getting 1557 descriptors for a system is given at: https://github.com/usnistgov/jarvis/blob/master/jarvis/sklearn/examples/desc_example.py

• Access to JARVIS database:

  Our database is freely available at https://www.ctcms.nist.gov/~knc6/JVASP.html, https://www.ctcms.nist.gov/jarvisml/, https://www.ctcms.nist.gov/~knc6/periodic.html, and https://www.ctcms.nist.gov/~knc6/JLAMMPS.html for JARVIS-DFT, JARVIS-ML and JARVIS-FF. We can also load the dataset using python scripts similar to https://github.com/knc6/jarvis/blob/master/jarvis/db/static/explore_db.py .

• Uploading your data using JARVIS-API:

  In addition to downloading/browsing through the JARVIS-database, one can also upload their data and query using JARVIS-API. Follow the instructions in https://github.com/usnistgov/jarvis/blob/master/jarvis/db/mdcs/mdcs_api.py

Founders

Kamal Choudhary, Francesca Tavazza (NIST)

Contributors

Daniel Wheeler, Faical Yannick Congo, Kevin Garrity, Brian DeCost, Adam Biacchi, Lucas Hale, Andrew Reid, Marcus Newrock (NIST)

Plot-gallery with additional jupyter notebooks

Notebook:	https://github.com/usnistgov/jarvis/blob/master/jarvis/db/static/RDF%2CPRDF%2CADF%2CDDF.ipynb

Notebook:	https://github.com/usnistgov/jarvis/blob/master/jarvis/db/static/RDF%2CPRDF%2CADF%2CDDF.ipynb

Notebook:	https://github.com/usnistgov/jarvis/blob/master/jarvis/db/static/RDF%2CPRDF%2CADF%2CDDF.ipynb

Notebook:	https://github.com/usnistgov/jarvis/blob/master/jarvis/db/static/RDF%2CPRDF%2CADF%2CDDF.ipynb

Notebook:	https://github.com/usnistgov/jarvis/blob/master/jarvis/db/static/band_structure.ipynb

Notebook:	https://github.com/usnistgov/jarvis/blob/master/jarvis/db/static/band_structure.ipynb

Notebook:	https://github.com/usnistgov/jarvis/blob/master/jarvis/db/static/Wulff.ipynb

Notebook:	https://github.com/usnistgov/jarvis/blob/master/jarvis/db/static/Boltztrap.ipynb

Notebook:	https://github.com/usnistgov/jarvis/blob/master/jarvis/db/static/Convergence.ipynb

Notebook:	https://github.com/usnistgov/jarvis/blob/master/jarvis/db/static/Convergence.ipynb