X-Git-Url: http://iramuteq.org/git?p=iramuteq;a=blobdiff_plain;f=textsimi.py;h=55e2f2b51f4d1055352151d9feb260cc7970a978;hp=1340276e9caf9507e420192515b5c4b2caf58b22;hb=f747a97a170eadbb78e7e69eeac27631ff55534c;hpb=8fa853a25a9d62b1446e1bc543e5a3a4d0e03dcf diff --git a/textsimi.py b/textsimi.py index 1340276..55e2f2b 100644 --- a/textsimi.py +++ b/textsimi.py @@ -1,32 +1,73 @@ # -*- coding: utf-8 -*- #Author: Pierre Ratinaud -#Copyright (c) 2008-2011 Pierre Ratinaud -#Lisense: GNU/GPL +#Copyright (c) 2008-2013 Pierre Ratinaud +#License: GNU/GPL -from chemins import ConstructPathOut, construct_simipath -from corpus import Corpus +from chemins import ffr, simipath import os -from ConfigParser import RawConfigParser -from guifunct import getPage, getCorpus -from dialog import StatDialog -from functions import indices_simi, progressbar, treat_var_mod -from tableau import Tableau -from tabsimi import DoSimi +from analysetxt import AnalyseText +from guifunct import PrepSimi +from functions import indices_simi, progressbar, treat_var_mod, read_list_file, print_liste +from PrintRScript import PrintSimiScript import wx from copy import copy +import logging -class SimiTxt : - def __init__(self, parent, cmd = False, param = None): - self.parent = parent - self.cmd = cmd - self.ConfigPath = parent.ConfigPath - self.DictPath = parent.DictPath - self.KeyConf = RawConfigParser() - self.KeyConf.read(self.ConfigPath['key']) +log = logging.getLogger('iramuteq.textsimi') + +class SimiTxt(AnalyseText): + def doanalyse(self) : + self.parametres['type'] = 'simitxt' + self.pathout.basefiles(simipath) self.indices = indices_simi + if self.dlg : + self.makesimiparam() + #FIXME + self.actives = self.corpus.make_actives_limit(3) + dictcol = dict([[i, [act, self.corpus.getlemeff(act)]] for i, act in enumerate(self.actives)]) + continu = False + if self.dlg : + self.listet = self.corpus.make_etoiles() + self.listet.sort() + self.stars = copy(self.listet) + self.parametres['stars'] = copy(self.listet) + self.parametres['sfromchi'] = False + self.dlg.Destroy() + prep = PrepSimi(self.ira, self, self.parametres, self.pathout['selected.csv'], self.actives, indices_simi, wordlist=dictcol) + if prep.val == wx.ID_OK : + continu = True + self.parametres = prep.parametres + self.dlg = progressbar(self.ira, 4) + else : + return False + else : + continu = True + if continu : + self.makefiles() + script = PrintSimiScript(self) + script.make_script() + if not self.doR(script.scriptout, dlg = self.dlg, message = 'R...') : + log.info('Problem') + return False + if self.parametres['type_graph'] == 1: + if self.parametres['svg'] : + filename, ext = os.path.splitext(script.filename) + fileout = filename + '.svg' + else : + fileout = script.filename + if os.path.exists(self.pathout['liste_graph']): + graph_simi = read_list_file(self.pathout['liste_graph']) + graph_simi.append([os.path.basename(fileout), script.txtgraph]) + else : + graph_simi = [[os.path.basename(fileout), script.txtgraph]] + print_liste(self.pathout['liste_graph'], graph_simi) + else : + return False + + def makesimiparam(self) : self.paramsimi = {'coeff' : 0, 'layout' : 2, - 'type' : 1, + 'type_graph' : 1, 'arbremax' : 1, 'coeff_tv' : 1, 'coeff_tv_nb' : 0, @@ -51,120 +92,93 @@ class SimiTxt : 'height' : 1000, 'bystar' : False, 'first' : True, - 'keep_coord' : True, + 'keep_coord' : False, 'alpha' : 20, 'film': False, + 'svg' : 0, + 'com' : 0, + 'communities' : 0, + 'halo' : 0, + #'ira' : self.pathout['Analyse.ira'] } - page = getPage(self.parent) - if page is not None : - self.corpus = getCorpus(page) - if self.corpus is not None : - self.pathout = ConstructPathOut(self.corpus.parametre['openpath'], 'simitxt') - self.dictpathout = construct_simipath(self.pathout) - self.val = wx.ID_OK - self.make_table() - self.make_simi() - else : - self.corpus = Corpus(parent) - self.corpus.content = self.parent.content - self.corpus.parametre['encodage'] = parent.corpus_encodage - self.corpus.parametre['lang'] = parent.corpus_lang - self.corpus.parametre['filename'] = parent.filename - self.corpus.parametre['eff_min_uce'] = None - self.pathout = ConstructPathOut(self.corpus.parametre['filename'], 'simitxt') - self.dictpathout = construct_simipath(self.pathout) - dial = StatDialog(self, self.parent) - dial.check_uce.SetValue(True) - dial.check_uce.Enable(False) - dial.OnCheckUce(wx.EVT_MENU) - self.val = dial.ShowModal() - if self.val == wx.ID_OK : - with open(self.parent.ConfigPath['key'], 'w') as f: - self.KeyConf.write(f) - if dial.radio_lem.GetSelection() == 0 : lem = True - else : lem = False - if dial.exp.GetSelection() == 0 : exp = True - else : exp = False - dial.Destroy() - self.corpus.parametre['lem'] = lem - self.corpus.parametre['expressions'] = exp - self.corpus.parametre['nbforme_uce'] = dial.spin_ctrl_4.GetValue() - self.corpus.parametre['eff_min_forme'] = dial.spin_ctrl_5.GetValue() - self.corpus.parametre['max_actives'] = dial.spin_max_actives.GetValue() - self.make_corpus() - self.make_table() - self.make_simi() - else : - dial.Destroy() + self.parametres.update(self.paramsimi) - def make_corpus(self) : - print 'make corpus' - if not self.cmd : - dlg = progressbar(self, maxi = 6) - self.corpus.supplementaires = [option for option in self.KeyConf.options('KEYS') if self.KeyConf.get('KEYS', option) == "2"] - self.corpus.typeactive = [option for option in self.KeyConf.options('KEYS') if self.KeyConf.get('KEYS', option) == "1"] - ucis_txt, ucis_paras_txt = self.corpus.start_analyse(self.parent, dlg = dlg, cmd = self.cmd) - del ucis_txt - - if not self.cmd : - dlg.Update(5, '%i ucis - Construction des uces' % len(ucis_paras_txt)) - self.corpus.make_ucis_paras_uces(ucis_paras_txt, make_uce = True) - del ucis_paras_txt - - if self.corpus.para_coords != [[] for val in self.corpus.para_coords] : - self.corpus.parametre['para'] = True - else : - self.corpus.parametre['para'] = False - self.corpus.make_etoiles(self.corpus.para_coords) - print 'len(ucis_paras_uces)', len(self.corpus.ucis_paras_uces) + def makefiles(self, lim=3) : + #self.actives, lim = self.corpus.make_actives_nb(self.parametres.get('max_actives',1500), 1) + self.parametres['eff_min_forme'] = lim + self.parametres['nbactives'] = len(self.actives) + self.parametres['fromprof'] = False + self.corpus.make_and_write_sparse_matrix_from_uces(self.actives, self.pathout['mat01.csv'], self.pathout['listeuce1.csv']) + with open(self.pathout['actives.csv'], 'w') as f : + f.write('\n'.join(self.actives).encode(self.ira.syscoding)) - if not self.cmd : - dlg.Update(6, u'Dictionnaires') - uces, self.orderuces = self.corpus.make_forms_and_uces() - self.corpus.ucenb = len(uces) - self.corpus.make_lems(self.parent.lexique) +class SimiFromCluster(SimiTxt) : + def __init__(self, ira, corpus, actives, lfreq, lchi, numcluster, parametres = None, dlg = False) : + self.actives = actives + self.numcluster = numcluster + self.lfreq = lfreq + self.lchi = lchi + parametres['name'] = 'simi_classe_%i' % (numcluster + 1) + dlg.Destroy() + SimiTxt.__init__(self, ira, corpus, parametres, dlg=True, lemdial = False) + + def preferences(self) : + return self.parametres - self.corpus.make_var_actives() - self.corpus.make_var_supp() - self.corpus.lems_eff = self.corpus.make_lem_eff() + def doanalyse(self) : + self.parametres['type'] = 'clustersimitxt' + self.pathout.basefiles(simipath) + self.indices = indices_simi + if self.dlg : + self.makesimiparam() + if 'bystar' in self.parametres : + del self.parametres['bystar'] + dictcol = dict([[i, [act, self.corpus.getlemclustereff(act, self.numcluster)]] for i, act in enumerate(self.actives)]) + continu = True + #print self.dlg + if self.dlg : + self.dlg.Destroy() + self.stars = []#copy(self.listet) + self.parametres['stars'] = 0#copy(self.listet) + self.parametres['sfromchi'] = 1 + prep = PrepSimi(self.ira, self, self.parametres, self.pathout['selected.csv'], self.actives, indices_simi, wordlist=dictcol) + if prep.val == wx.ID_OK : + continu = True + self.parametres = prep.parametres + else : + continu = False + if continu : + self.dlg = progressbar(self.parent, 3) + self.makefiles() + self.parametres['type'] = 'clustersimitxt' + script = PrintSimiScript(self) + script.make_script() + if not self.doR(script.scriptout, dlg = self.dlg, message = 'R ...') : + return False + if self.parametres['type_graph'] == 1: + if self.parametres['svg'] : + filename, ext = os.path.splitext(script.filename) + fileout = filename + '.svg' + else : + fileout = script.filename + if os.path.exists(self.pathout['liste_graph']): + graph_simi = read_list_file(self.pathout['liste_graph']) + graph_simi.append([os.path.basename(fileout), script.txtgraph]) + else : + graph_simi = [[os.path.basename(fileout), script.txtgraph]] + print_liste(self.pathout['liste_graph'], graph_simi) + else : + return False - #variables = treat_var_mod(listet) - #print(variables) - #self.corpus.write_etoiles(self.dictpathout['etoiles']) - if not self.cmd : - dlg.Destroy() - - def make_table(self) : - if 'orderuces' not in dir(self) : - self.orderuces = [(i,j,k) for i, uci in enumerate(self.corpus.ucis_paras_uces) for j, para in enumerate(uci) for k, uce in enumerate(para)] - self.orderuces = dict([[val, i] for i, val in enumerate(self.orderuces)]) - self.corpus.ucenb = len(self.orderuces) - #tabuc1 = self.corpus.make_table_with_uce(self.orderuces) - #tabuc1.insert(0,self.corpus.actives) - #tabuc1 = self.corpus.make_sparse_matrix_with_uce(self.orderuces) - #self.corpus.write_sparse_matrix(self.dictpathout['mat01'], tabuc1, self.corpus.ucenb, len(self.corpus.actives)) - if self.corpus.actives is None : - self.corpus.typeactive = [option for option in self.KeyConf.options('KEYS') if self.KeyConf.get('KEYS', option) == "1"] - self.corpus.min_eff_formes() - self.corpus.make_var_actives() - self.corpus.make_and_write_sparse_matrix_from_uce(self.orderuces, self.dictpathout['mat01']) - #self.corpus.write_tab(tabuc1,self.dictpathout['mat01']) + def makefiles(self) : + self.parametres['eff_min_forme'] = 3 + self.parametres['nbactives'] = len(self.actives) + self.parametres['fromprof'] = True + self.corpus.make_and_write_sparse_matrix_from_classe(self.actives, self.corpus.lc[self.numcluster], self.pathout['mat01.csv']) + with open(self.pathout['actives.csv'], 'w') as f : + f.write('\n'.join(self.actives).encode(self.ira.syscoding)) + with open(self.pathout['actives_nb.csv'], 'w') as f : + f.write('\n'.join([`val` for val in self.lfreq])) + with open(self.pathout['actives_chi.csv'], 'w') as f : + f.write('\n'.join([`val` for val in self.lchi])) - def make_simi(self) : - self.tableau = Tableau(self.parent, '') - self.tableau.listactives = self.corpus.actives - self.tableau.parametre['fromtxt'] = True - if 'lems_eff' not in dir(self.corpus) : - self.corpus.lems_eff = self.corpus.make_lem_eff() - #print('ATTENTION ETOILES') - #self.paramsimi['bystar'] = True - self.listet = self.corpus.get_unique_etoiles() - self.listet.sort() - self.tableau.listet = copy(self.listet) - self.paramsimi['stars'] = copy(self.listet) - #self.paramsimi['cexfromchi'] = True - self.paramsimi['sfromchi'] = False - #self.paramsimi['vlabcolor'] = True - self.tableau.actives = dict([[lem, self.corpus.lems_eff[lem]] for lem in self.corpus.actives]) - self.corpus.save_corpus(self.dictpathout['corpus']) - DoSimi(self, fromprof = self.dictpathout['mat01'], param = self.paramsimi, pathout = self.pathout)