I have two images from a Monte Carlo Basin Hopping where silicon (in yellow) is getting replaced by germanium (in blue). Both have 32 germanium atoms at Si/Ge = 2.5; only the Ge distribution is changing.
There appears to be a clustering of the germanium in the the first structure with relative energy= 0 ev. Whereas the higher energy structure has no such pattern. I am trying to to correlate the energy and the clustering.
How could I verify this layering/clustering, to correlate the energy and the clustering, using some sort of metrics?
Any of the following would be helpful for me.
- Any links to any publication to gain ideas
- Suggestion of a metric or parameter.
- Any ML method or software that may help with this.
Thanks in advance.