cjasen
commented
2 weeks ago ¿Seguro que hay que normalizar dividiendo por (t-s+1)? Supongo que sí, ya que al calcular las
En calc_thermo (el archivo se llama calc_thermo_loopy):
Pero la verdad es que el resultado creo que tiene más sentido, porque para
f(L) is a function which supposly tell us the fraction of islands with length L. However, configurations with almost all their residues in state <m_ksigma_k>=0.999... return f(1)=1 (in theory, f(1)=1 would mean the <m_ksigma_k> state of residues like this 01010101). There's some issue with normalization, because the total area under the curve is (or at least can be) greater than one, but I don't think this is a serious issue.
I'd like to have a function where for all <m_k*sigma_k>=1 then f(L)=0 for all L but f(L=N)=1. With this formula, I think the result for this example would be f(L)=1 for all L. This isn't, however, the result (but maybe it isn't coded correctly?).
See here, for example, where for T=50 and 450 all <m_k*sigma_k>=~1: