Data and scripts for RDF recognition of nanoparticles architecture
Description
The ultimate sensitivity of pair radial distribution functions (RDF) of atoms to the nanoparticle architecture. Each Jupiter notebook is presented in two formats: interactive with extension `ipynb` and static, with extension `html`. Files included: 1. XAFS folder contanins X-ray absorption spectra (*.norm files) and extracted from them EXAFS oscillation functions (*.chi files) for the samples. Also added examples Larch script (*.lar) for fitting of the EXAFS data for sample PtCu_stage2 and example Bash script (*.sh) to produce a set of feff dat files. 1. `produce_rdfs.py` -- Python script to construct nanoparticle models and calculate RDFs for them. Requires ASE [https://wiki.fysik.dtu.dk/ase/ase] and bimetall [https://github.com/lavakyan/ase-bimetall] libraries. 1. `bimetall.tgz` -- bimetall library, requred for running of `produce_rdfs.py` script. 1. `rdfs.tgz` -- archive containing the RDFs obtained after run of `produce_rdfs.py` script. 1. `postproc_RDFs.ipynb` -- Jupyter notebook for loading, scaling and production of `*.pkl` data files. 1. `analyze_R1.ipynb` -- Jupyter notebook for analyzis of metal-metal interatomic distances. Gaussian fitting of RDFs. Classification using ML methods. Prediction of architectures for the samples. 1. `classify_RDFs.ipynb` -- Jupyter notebook for analysis of theoretical RDFs on a range 2 < R < 5 (three coordination shells) and predictions of nanoparticle architectures. 1. `classify_RDFs.ipynb` -- Jupyter notebook for analysis of theoretical RDFs on a range 2.2 < R < 3.2 (first coordination shell) and predictions of nanoparticle architectures.