view article

Figure 1
Pseudo-code describing how to create POM clusters from a CIF database and how to train POMFinder. A POM database is built from the ICSD and COD by cutting out clusters from all crystal structures with chemical compositions similar to POM clusters in solution (step 1) (Gumerova & Rompel, 2020BB31). A number of PDFs are simulated for each POM cluster using various parameters (step 2). These PDFs are then used to train a GBDT model for classifying the corresponding structure from a PDF (step 3).

Journal logoJOURNAL OF
APPLIED
CRYSTALLOGRAPHY
ISSN: 1600-5767
Follow J. Appl. Cryst.
Sign up for e-alerts
Follow J. Appl. Cryst. on Twitter
Follow us on facebook
Sign up for RSS feeds