From b1508377fa575a61f1127273d328766608237014 Mon Sep 17 00:00:00 2001 From: Pierre Date: Tue, 11 Feb 2014 14:13:32 +0100 Subject: [PATCH] add banalites --- textaslexico.py | 21 ++++++++++++++++++--- 1 file changed, 18 insertions(+), 3 deletions(-) diff --git a/textaslexico.py b/textaslexico.py index 194c4dd..1f3e32f 100644 --- a/textaslexico.py +++ b/textaslexico.py @@ -1,7 +1,7 @@ # -*- coding: utf-8 -*- #Author: Pierre Ratinaud #Copyright (c) 2008-2011 Pierre Ratinaud -#Lisense: GNU/GPL +#License: GNU/GPL from chemins import ConstructPathOut, StatTxtPathOut #from corpus import Corpus @@ -11,7 +11,7 @@ import os import sys from listlex import * from functions import exec_rcode, progressbar, check_Rresult, CreateIraFile, print_liste, treat_var_mod, write_tab, DoConf -from dialog import OptLexi, StatDialog #LexDialog +from dialog import OptLexi, StatDialog from openanalyse import OpenAnalyse import tempfile from ConfigParser import RawConfigParser @@ -46,6 +46,9 @@ class Lexico(AnalyseText) : txt += """ dmt<-read.csv2("%s",row.names=1) """ % self.dictpathout['tabletypem'] + txt += """ + indice <- "%s" + """ % self.parametres['indice'] if self.parametres['indice'] == 'hypergeo' : txt += """ outf <- make.spec.hypergeo(dmf) @@ -56,7 +59,19 @@ class Lexico(AnalyseText) : outf<-AsLexico2(dmf) outt<-AsLexico2(dmt) """ - + txt += """ + if (indice == 'hypergeo') { + banseuil <- 2 + } else if (indice == 'chi2') { + banseuil <- 3 + } + banal <- apply(abs(outf[[1]]), 1, max) + banal <- which(banal < banseuil) + banalfreq <- rowSums(dmf[banal,]) + banalspec <- specf<-outf[[1]][banal,] + banal <- cbind(banalfreq, banalspec) + write.csv2(banal,file="%s") + """ % self.pathout['banalites.csv'] txt += """ specf<-outf[[1]] spect<-outt[[1]] -- 2.7.4