X-Git-Url: http://iramuteq.org/git?p=iramuteq;a=blobdiff_plain;f=corpus.py;h=e222940b42281882adcfc9828b4b2c1c012325fc;hp=76ebb660911223bf31496b8d0a23aea56c4d2290;hb=dbc9195121789297412a7d46eb9c1f089b77490e;hpb=432118f2ac3d2f8234c388e77d0fb9e14234750f

diff --git a/corpus.py b/corpus.py
index 76ebb66..e222940 100644
--- a/corpus.py
+++ b/corpus.py
@@ -8,7 +8,7 @@ _ = gettext.gettext
 import locale
 import sys
 from time import time
-from functions import decoupercharact, ReadDicoAsDico, DoConf
+from functions import decoupercharact, ReadDicoAsDico, DoConf, ReadLexique
 import re
 import sqlite3
 import itertools
@@ -17,7 +17,6 @@ from operator import itemgetter
 from uuid import uuid4
 from chemins import PathOut
 from dialog import CorpusPref
-from functions import ReadLexique, ReadDicoAsDico
 from colors import colors
 import datetime
 
@@ -38,8 +37,7 @@ def copycorpus(corpus) :
 
 class Corpus :
     """Corpus class
-    list of uci
-
+    list of text
     """
     def __init__(self, parent, parametres = {}, read = False) :
         self.parent = parent
@@ -77,10 +75,10 @@ class Corpus :
                 gramtype = self.parent.lexique[word][1]
                 lem = self.parent.lexique[word][0]
             elif word.isdigit() :
-                gramtype = 'num'
+                gramtype = u'num'
                 lem = word
             else :
-                gramtype = 'nr'
+                gramtype = u'nr'
                 lem = word
             self.formes[word] =  Word(word, gramtype, len(self.formes), lem)
             self.idformesuces[self.formes[word].ident] = {self.ucis[-1].uces[-1].ident : 1}
@@ -275,8 +273,12 @@ class Corpus :
         if self.iduces is None :
             self.iduces = dict([[uce.ident, uce] for uci in self.ucis for uce in uci.uces])
 
-    def make_lexitable(self, mineff, etoiles) :
-        tokeep = [lem for lem in self.lems if self.lems[lem].freq >= mineff]
+    def make_lexitable(self, mineff, etoiles, gram = 0) :
+        if gram == 0 :
+            grams = {1:'', 2:''}
+        else :
+            grams = {gram :''}
+        tokeep = [lem for lem in self.lems if self.lems[lem].freq >= mineff and self.lems[lem].act in grams]
         etuces = [[] for et in etoiles]
         for uci in self.ucis :
             get = list(set(uci.etoiles).intersection(etoiles))
@@ -434,7 +436,7 @@ class Corpus :
                 f.write(guce.encode(self.parametres['syscoding']) + '\n\n')
 
     def export_classe(self, outf, classe, lem = False) :
-        sts = self.lc[classe] 
+        sts = self.lc[classe - 1] 
         res = self.getconcorde(sts)
         self.make_iduces()
         with open(outf, 'w') as f :
@@ -506,6 +508,17 @@ class Corpus :
                 table_uce[uces[uce]][i] = 1
         table_uce.insert(0, list_act)
         return table_uce      
+    
+    def make_pondtable_with_classe(self, uces, list_act) :
+        table_uce = [[0 for val in list_act] for line in range(0,len(uces))]
+        uces = dict([[uce, i] for i, uce in enumerate(uces)])
+        for i, lem in enumerate(list_act) :
+            uceseff = self.getlemuceseff(lem)
+            lemuces = list(set(uceseff.keys()).intersection(uces))
+            for uce in lemuces :
+                table_uce[uces[uce]][i] = uceseff[uce]
+        table_uce.insert(0, list_act)
+        return table_uce 
 
     def parse_active(self, gramact, gramsup = None) :
         log.info('parse actives')
@@ -514,7 +527,7 @@ class Corpus :
                 self.lems[lem].act = 2
             elif self.lems[lem].gram in gramact :
                 self.lems[lem].act = 1
-            elif gramsup is not None :
+            elif gramsup is not None and self.lems[lem].gram not in gramact:
                 if self.lems[lem].gram in gramsup :
                     self.lems[lem].act = 2
                 else :
@@ -534,6 +547,8 @@ class Corpus :
         allactives = [[self.lems[lem].freq, lem] for lem in self.lems if self.lems[lem].act == key and self.lems[lem].freq >= 3]
         self.activenb = len(allactives)
         allactives = sorted(allactives, reverse = True)
+        if self.activenb == 0 :
+            return [], 0
         if len(allactives) <= nbmax :
             log.info('nb = %i - eff min = %i ' % (len(allactives), allactives[-1][0]))
             return [val[1] for val in allactives], allactives[-1][0]
@@ -830,8 +845,7 @@ class MakeUciStat :
         ucinb = corpus.getucinb()
         ucisize = corpus.getucisize()
         ucimean = float(sum(ucisize))/float(ucinb)
-        detoile = corpus.make_etoiles_dict()
-        
+        detoile = corpus.make_etoiles_dict() 
 
 class Uci :
     def __init__(self, iduci, line, paraset = None) :
@@ -1081,7 +1095,7 @@ class BuildCorpus :
     def firstclean(self, txt) :
         txt = txt.replace(u'’',"'")
         txt = txt.replace(u'œ', u'oe')
-        return txt.replace('...',u' £$£ ').replace('?',' ? ').replace('.',' . ').replace('!', ' ! ').replace(',',' , ').replace(';', ' ; ').replace(':',' : ').replace(u'…', ' £$£ ')
+        return txt.replace('...',u' £$£ ').replace('?',' ? ').replace('.',' . ').replace('!', ' ! ').replace(',',' , ').replace(';', ' ; ').replace(':',' : ').replace(u'…', u' £$£ ')
 
     def make_cleans(self, txt) :
         for clean in self.cleans :