5. Development of cluster expansion model

1. Collecting DFT data

   # End members are identified with --end option.
   # --end (location of vasprun.xml)
   #       (number of formula unit in structure specified by vasprun.xml)
   #       (element type1) (element type2) ...
   
   $(pyclupan)/tools/collect_dft.py --end finished/1-0-0/vasprun.xml 2.0 0 2 --end finished/1-0-3/vasprun.xml 2.0 1 3 --vaspruns finished/*/vasprun.xml
   

2. Model estimation

   $(pyclupan)/regression/regression.py -x ../4-correlation/correlations.pkl -y ../2-dft/summary_dft.yaml --upper_bound 0.5
   

3. Prediction by cluster expansion model

   # Correlation functions for target structures must be computed in advance.
   
   $(pyclupan)/prediction/prediction.py -x ../4-correlation/correlations.pkl --coeffs ../5-reg/regression.yaml