Double check epsilon implementation and implement constants and weights for epsilon and nu in the import function rather than in each script

[giovannipizzi]

Discussed the test of epsilon with @eimrek

[eimrek]

Hi, just to have a reply on this issue:

I tested epsilon and nu for "Fr2O5" directly in https://github.com/aiidateam/acwf-verification-scripts/blob/main/acwf_paper_plots/plots/supplementary/periodic_tables_all/generate_all_periodic_table.py

and they matched with my javascript implementation (using gaussian quadrature), given here: https://github.com/materialscloud-org/acwf-verification/blob/main/src/components/ComparisonMeasures.js

(numbers also available at https://acwf-verification.materialscloud.org)

Should there be any more thorough testing?