Quantum Monte Carlo Database (QMC-DB)
- Raymond Plante
Project Lead: Elif Ertekin (ertekin at illinois.edu)
Project Aims
We look to create a community-focused database of results and experience employing the Quantum Monte Carlo (QMC) computational technique for modeling materials. It leverages our recent Blue Waters award in which we will apply this technique to create a large corpus of models. We are building a framework for capturing input models and parameters and automating the launching of simulations on Blue Waters. In this project, we want to merge this information with the computational outputs and derived material properties and load them into a searchable database. In order to share this database with the larger materials community, we plan to build a web-based portal to the database that allows scientists to explore results.
A key aim for sharing this data is to help the computational materials community to build greater intuition on the use of the QMC technique. Although the QMC technique is one of the most accurate electronic structure methods available, the application of the technique to real materials is still relatively new. With further funding (via proposals to be developed this fall), we hope to allow other groups to add their results and insights to the portal to further build the community knowledge base. This project should be useful to the development of the NDS as it will provide insight and an example of integrating data that is not simply file-based. We hope it will also provide opportunities for integration with the Materials Data Facility.
Software tools and data resources being leverage
Exising Tools to be Leveraged
- MongoDB – for capturing computational inputs and outputs
- JSON – format for data exchange between computational runs and the database
- (some web development framework)
- (data storage management tool)
New Data Collections
The project will host results from QMC simulations, primarily via the MongoDB database.
Project Software Repository
System software will be hosted in !GitHub.
- Simulation software and automation: QWalk
- QMCDB system software on !GitHub: QWalk/QMCDB
Project Team Members
- Lucas Wagner (UIUC)
- Ray Plante (NIST)
- Josh Schiller (UIUC)
More Information
- References:
- MongoDB Manual
- Note that this has pointers to different versions of the "Getting Started Guide", including one based on the shell client and another based on the PyMongo python library.
- MongoDB Manual
- Ray's Notes...
- ...on using Labs (basis for documentation under development)
- ...on MongoDB
- Development How-Tos
Project Status
Project Start: 14 Sept. 2015
- 3-month status update
- 6-month status update
07 Oct 2015
We have an instance of MongoDB running in NDS Labs so that we can get familiar with the interfaces and experiment with our document schemas.
14 Sept 2015
Accounts are being set up, and we're getting familiar with the nebular system.