Automated Data Analysis
NREL has developed both local and network-based analysis capabilities to more efficiently turn data from high-throughput experiments into useful materials knowledge.
Data generated during combinatorial depositions and subsequent spatially resolved characterization of combinatorial libraries is of tremendous value.
![mage showing automated management and analysis process: (a) network, (b) analysis, (c) web, (d) extract, (e) transform, and (f) load](/materials-science/assets/images/automated-data-analysis-data-management.png)
Significant investments have been made to support the navigation of this data stream by researchers, they include:
- Local computer analysis – custom software package COMBIgor installed on user computers implemented in Igor-PRO, which offers project level management and analysis of data from combinatorial experiments.
- Network accessible database – interactive visualization of deposition and characterization data from combinatorial experiments in High-Throughput Experimental Materials Database (HTEM-DB)
- Research data Infrastructure – automated data harvesting, processing and alignment of deposition and characterization data from combinatorial experiments.
COMBIgor
![Chart showing COMBIgor data](/materials-science/assets/images/automated-data-analysis-combigor.png)
COMBIgor is an open-source software package written in the Igor-PRO environment and designed to offer a systematic approach to loading, storing, processing, and visualizing combinatorial data. Some features include:
- Methods for loading and storing data sets from combinatorial libraries
- Data-analysis routines for simplified processing across libraries of data
- Advanced visualization routines to construct informative figures
- Extensive documentation and example code for easy user customization.
COMBIgor is designed to be easily customized for integration with additional instruments and data processing algorithms.
For more information, read COMBIgor: Data Analysis Package for Combinatorial Materials Science, ACS Comb. Sci. (2019).
High-Throughput Experimental Materials Database
![Data visualization from the HTEM-DB](/materials-science/assets/images/automated-data-analysis-htem-db.jpg)
The HTEM-DB contains the materials data obtained from high-throughput experiments.
HTEM-DB was designed with the purpose of releasing large amounts of high-quality experimental data to the public through an interactively searchable and programmatically accessible long-term repository. Some features include:
- Web-based interactive search and filter tools for finding entries of interest
- Data visualization tools for inspection and comparison of data sets
- Application programming interface with machine learning algorithms
- More than 100,000 unique compositions at unique processing conditions.
For more information, read An Open Experimental Database for Exploring Inorganic Materials, Scientific Data (2018).
Research Data Infrastructure
The HTEM-DB is built from an extensive research data infrastructure for high-throughput experimental materials science. It features a data warehouse, which is a raw file repository automatically harvested from laboratory computers and stored in a central location, and a custom web form-based application for capturing synthesis conditions associated with the characterized sample libraries known as the Lab Metadata Collector. Some features include:
- Digital data collection by a data warehouse
- Metadata collection by the Laboratory Metadata Collector
- Data extraction, transformation, and loading
- Direct links to HTEM-DB and COMBIgor.
For more information, read Research Data Infrastructure for High-Throughput Experimental Materials, Patterns (2021).
Publications
An Open Experimental Database for Exploring Inorganic Materials, Scientific Data (2018)
COMBIgor: Data Analysis Package for Combinatorial Materials Science, ACS Combinatorial Science (2019)
Research Data Infrastructure for High-Throughput Experimental Materials, Patterns (2021)
Contacts
Share