import re
import matplotlib.pyplot as plt
import numpy as np
with open('testi/collodi_pinocchio.txt', mode='r', encoding='iso-8859-1') as f: 
	testo = f.read()
testo=testo[1704:]
testo=re.sub("'","' ", testo)
testo=re.sub('[\.,;:\-?!]{1}','',testo)
testo=testo.lower()
testo=testo.split()
fd=[testo.count(i) for i in set(testo)]
fd.sort(reverse=True)
V=len(set(testo))
fig=plt.figure()
ax=fig.add_subplot(1,1,1)
ax.set_xscale('log')
ax.set_yscale('log')
ax.plot(range(1,V+1),fd, 'b.')
x=np.arange(1,V+1)
y=5850.7/(x**1.03372)
ax.plot(x,y, 'r-')
plt.show()