+
+ def find_segments_in_classe(self, list_uce, taille_segment, taille_limite):
+ d={}
+ for uce in self.getconcorde(list_uce) :
+ uce = uce[1].split()
+ d =self.count_from_list([' '.join(uce[i:i+taille_segment]) for i in range(len(uce)-(taille_segment - 1))], d)
+ l = [[d[val], val, taille_segment] for val in d if d[val] >= 3]
+ del(d)
+ l.sort()
+ if len(l) > taille_limite :
+ l = l[-taille_limite:]
+ return l
+
+ def make_segments_profile(self, fileout, lenmin = 3, lenmax = 10, effmin = 50, lem = False) :
+ d = {}
+ for b, classe in enumerate(self.lc) :
+ for uce in self.getconcorde(classe) :
+ uce = uce[1].split()
+ if lem :
+ uce = [self.formes[forme].lem for forme in uce]
+ for taille_segment in range(lenmin,lenmax) :
+ d =self.count_from_list_cl([' '.join(uce[i:i+taille_segment]) for i in range(len(uce)-(taille_segment - 1))], d, b, len(self.lc))
+ result = [[seg] + [str(val) for val in d[seg]] for seg in d if sum(d[seg]) >= effmin]
+ with open(fileout, 'w') as f :
+ f.write('\n'.join([';'.join(line) for line in result]))
+
+ def make_proftype(self, outf) :
+ res = {}
+ for lem in self.lems :
+ gram = self.lems[lem].gram
+ if not gram in res :
+ res[gram] = [0 for val in self.lc]
+ lemuceeff = self.getlemuceseff(lem)
+ for i, classe in enumerate(self.lc) :
+ concern = set(classe).intersection(lemuceeff.keys())
+ res[gram][i] += sum([lemuceeff[uce] for uce in concern])
+ res = [[gram] + [`val` for val in res[gram]] for gram in res]
+ res.sort()
+ with open(outf, 'w') as f :
+ f.write('\n'.join([';'.join(line) for line in res]).encode(self.parametres['syscoding']))
+
+