EELSLab: a Python toolbox for (hyper)spectroscopy data analysis
F. de la Peña1, 3, *, M. Sarahan2, S. Mazzucco4, 5, L-F. Zagonel1, **, M. Walls1
1) Laboratoire de Physique des Solides, Bât. 510, Université Paris-Sud. 91405 Orsay Cedex
2) SuperSTEM, STFC Daresbury Laboratories. Keckwick Lane, Warrington WA4 4AD – UK
3) CEA-LETI, MINATEC 17, avenue des Martyrs, 38054 GRENOBLE Cedex 9 – France
4) Center for Nanoscale Science and Technology, National Institute of Science and Technology, 100 Bureau Drive, Gaithersburg, MD 20899-6203, USA
5) Institute for Research in Electronics and Applied Physics (Bldg. 223), Paint Branch Drive, University of Maryland, College Park, MD 20742-3511, USA
* Current address: department of Materials Science and Metallurgy, University of Cambridge. Pembroke Street, Cambridge, CB2 3QZ – UK
** Current address: Associação Brasileira de Tecnologia de Luz Sincrotron, Laboratório de Microscopia Eletrônica.13083-970 - Campinas, SP – Brasil
Modern scientific instruments from several disciplines now yield multidimensional spectroscopic data. As an example, a modern transmission electron microscope (TEM) can acquire spectral data from sub-atomic volumes that ultimately could reveal the position and nature of each atom of a material . However, such datasets are usually quite large and difficult to work with.
EELSLab  has been developed as a tool to facilitate hyperspectral data analysis. Originally it was intended for TEM data analysis, but it has been successfully used in other domains. Specifically, it provides easy access to multidimensional curve fitting, peak analysis and machine learning algorithms, as well as a viewing framework for navigating data and reading and writing capabilities for several popular hyperspectral formats.
This talk will discuss how Python has been used to implement these features, with demonstrations of applications to both spatially-resolved spectroscopic data (so-called spectrum images)  and to structural analysis of atomic resolution image stacks . The blend of intuitiveness, power, and availability of high-quality scientific libraries that Python offers has allowed the creation of a simple, natural tool that scientists from many disciplines can both use and easily extend into new scientific domains.
 Sandra Van Aert et al., « Three-dimensional atomic imaging of crystalline nanoparticles », Nature 470, no. 7334 (17th of February, 2011): 374-377.
 R. Arenal et al., « Extending the analysis of EELS spectrum-imaging data, from elemental to bond mapping in complex nanostructures », Ultramicroscopy 109, no. 1 (December 2008): 32-38.
 Michael C. Sarahan et al., « Point defect characterization in HAADF-STEM images using multivariate statistical analysis », Ultramicroscopy 111, no. 3 (February 2011): 251-257.