Metric for linear and 0.5 cases

If I try to compute more tan 10^7 permutations, Matlab runs out of memory. This can be solved, by saving the results in files as they are computed. But for now, I post results for 10^7 permutations (run in remotón).

Calculations for alfa, beta of our pnas, exponent 1:

pos_opt=coste2pos_num_ruido(todas.A*.05,todas.M*1.5+todas.S,todas.f,1,0);
[m,coste_real,coste_perm]=metrica_pot(todas.A*.05,todas.M*1.5+todas.S,todas.f,todas.pos_real,1,10^7,pos_opt,4);
hist(coste_perm,100)
>> hold on
>> ejes=axis;
>> plot(coste_real*[1 1],ejes(3:4),’k’)
>> save metrica_alfabetapnas_lineal m coste_perm coste_real

m is 0, no permutation has lower cost than the real case. So p<10^-7.

Calculations for alfa and beta predicted by Bayes, and exponent 0.5:

>> [m,coste_real,coste_perm]=metrica_pot(todas.A*10^-1.1,todas.M*10^-.9+todas.S,todas.f,todas.pos_real,0.5,10^7,datos.pos_opt,4);
>> cd ..
>> cd metricaelegans
>> save metrica_alfabetabayes_cerocinco m coste_perm coste_real
>> m
m =
0
>> close all
>> hist(coste_perm,100)
>> hold on
>> ejes=axis;
>> plot(coste_real*[1 1],ejes(3:4),’k’)