X-Git-Url: http://iramuteq.org/git?p=iramuteq;a=blobdiff_plain;f=PrintRScript.py;h=deaddf51ced131c513700e1eda655a5e86b25a58;hp=4d0fb40fcaad3634a92fd3ad88a1869a9d0234e4;hb=3d1a621b481e251f8a086af586c7eb0bb87b0004;hpb=8fa853a25a9d62b1446e1bc543e5a3a4d0e03dcf diff --git a/PrintRScript.py b/PrintRScript.py index 4d0fb40..deaddf5 100644 --- a/PrintRScript.py +++ b/PrintRScript.py @@ -1,7 +1,7 @@ # -*- coding: utf-8 -*- #Author: Pierre Ratinaud #Copyright (c) 2008-2011 Pierre Ratinaud -#Lisense: GNU/GPL +#License: GNU/GPL import tempfile from chemins import ffr @@ -17,8 +17,9 @@ class PrintRScript : log.info('Rscript') self.pathout = analyse.pathout self.analyse = analyse + self.parametres = analyse.parametres self.scriptout = self.pathout['temp'] - self.script = u"#Script genere par IRaMuTeQ - %s" % datetime.now().ctime() + self.script = u"#Script genere par IRaMuTeQ - %s\n" % datetime.now().ctime() def add(self, txt) : self.script = '\n'.join([self.script, txt]) @@ -32,11 +33,15 @@ class PrintRScript : def sources(self, lsources) : for source in lsources : - self.add('source("%s")' % source) + self.add('source("%s", encoding = \'utf8\')' % ffr(source)) + + def packages(self, lpks) : + for pk in lpks : + self.add('library(%s)' % pk) def load(self, l) : for val in l : - self.add('load("%s")' % val) + self.add('load("%s")' % ffr(val)) def write(self) : with open(self.scriptout, 'w') as f : @@ -46,6 +51,8 @@ class PrintRScript : class chdtxt(PrintRScript) : pass +def Rcolor(color) : + return str(color).replace(')', ', max=255)') class Alceste2(PrintRScript) : def doscript(self) : @@ -102,13 +109,13 @@ class Alceste2(PrintRScript) : # -def RchdTxt(DicoPath, RscriptPath, mincl, classif_mode, nbt = 9, libsvdc = False, libsvdc_path = None, R_max_mem = False): +def RchdTxt(DicoPath, RscriptPath, mincl, classif_mode, nbt = 9, svdmethod = 'svdR', libsvdc = False, libsvdc_path = None, R_max_mem = False, mode_patate = False): txt = """ source("%s") source("%s") source("%s") source("%s") - """ % (RscriptPath['CHD'], RscriptPath['chdtxt'], RscriptPath['anacor'], RscriptPath['Rgraph']) + """ % (ffr(RscriptPath['CHD']), ffr(RscriptPath['chdtxt']), ffr(RscriptPath['anacor']), ffr(RscriptPath['Rgraph'])) if R_max_mem : txt += """ memory.limit(%i) @@ -117,93 +124,112 @@ def RchdTxt(DicoPath, RscriptPath, mincl, classif_mode, nbt = 9, libsvdc = False txt += """ nbt <- %i """ % nbt - if libsvdc : + if svdmethod == 'svdlibc' and libsvdc : txt += """ - libsvdc <- TRUE + svd.method <- 'svdlibc' libsvdc.path <- "%s" """ % ffr(libsvdc_path) + elif svdmethod == 'irlba' : + txt += """ + library(irlba) + svd.method <- 'irlba' + libsvdc.path <- NULL + """ else : txt += """ - libsvdc <- FALSE + svd.method = 'svdR' libsvdc.path <- NULL """ - + if mode_patate : + txt += """ + mode.patate = TRUE + """ + else : + txt += """ + mode.patate = FALSE + """ txt +=""" library(Matrix) data1 <- readMM("%s") data1 <- as(data1, "dgCMatrix") row.names(data1) <- 1:nrow(data1) - """ % DicoPath['TableUc1'] + """ % ffr(DicoPath['TableUc1']) if classif_mode == 0: txt += """ data2 <- readMM("%s") data2 <- as(data2, "dgCMatrix") row.names(data2) <- 1:nrow(data2) - """ % DicoPath['TableUc2'] + """ % ffr(DicoPath['TableUc2']) txt += """ - chd1<-CHD(data1, x = nbt, libsvdc = libsvdc, libsvdc.path = libsvdc.path) + chd1<-CHD(data1, x = nbt, mode.patate = mode.patate, svd.method = svd.method, libsvdc.path = libsvdc.path) """ if classif_mode == 0: txt += """ - chd2<-CHD(data2, x = nbt, libsvdc = libsvdc, libsvdc.path = libsvdc.path) + chd2<-CHD(data2, x = nbt, mode.patate = mode.patate, svd.method = svd.method, libsvdc.path = libsvdc.path) """ - else: - txt += """ - chd2<-chd1 - """ txt += """ #lecture des uce listuce1<-read.csv2("%s") - """ % DicoPath['listeuce1'] + """ % ffr(DicoPath['listeuce1']) if classif_mode == 0: txt += """ listuce2<-read.csv2("%s") - """ % DicoPath['listeuce2'] + """ % ffr(DicoPath['listeuce2']) txt += """ -# rm(data1) + rm(data1) """ if classif_mode == 0: txt += """ -# rm(data2) + rm(data2) """ txt += """ - chd.result <- Rchdtxt("%s",mincl=%i,classif_mode=%i, nbt = nbt) + classif_mode <- %i + mincl <- %i + uceout <- "%s" + if (classif_mode == 0) { + chd.result <- Rchdtxt(uceout, chd1, chd2 = chd2, mincl = mincl,classif_mode = classif_mode, nbt = nbt) + } else { + chd.result <- Rchdtxt(uceout, chd1, chd2 = chd1, mincl = mincl,classif_mode = classif_mode, nbt = nbt) + } n1 <- chd.result$n1 classeuce1 <- chd.result$cuce1 - classeuce2 <- chd.result$cuce2 - """ % (DicoPath['uce'], mincl, classif_mode) + classes<-n1[,ncol(n1)] + write.csv2(n1, file="%s") + rm(n1) + """ % (classif_mode, mincl, ffr(DicoPath['uce']), ffr(DicoPath['n1.csv'])) txt += """ tree.tot1 <- make_tree_tot(chd1) # open_file_graph("%s", widt = 600, height=400) # plot(tree.tot1$tree.cl) # dev.off() - """%DicoPath['arbre1'] + """ % ffr(DicoPath['arbre1']) if classif_mode == 0: txt += """ + classeuce2 <- chd.result$cuce2 tree.tot2 <- make_tree_tot(chd2) # open_file_graph("%s", width = 600, height=400) # plot(tree.tot2$tree.cl) # dev.off() - """ % DicoPath['arbre2'] + """ % ffr(DicoPath['arbre2'] ) txt += """ tree.cut1 <- make_dendro_cut_tuple(tree.tot1$dendro_tuple, chd.result$coord_ok, classeuce1, 1, nbt) save(tree.cut1, file="%s") - classes<-n1[,ncol(n1)] + open_file_graph("%s", width = 600, height=400) - plot.dendropr(tree.cut1$tree.cl,classes) + plot.dendropr(tree.cut1$tree.cl,classes, histo=TRUE) open_file_graph("%s", width = 600, height=400) plot(tree.cut1$dendro_tot_cl) dev.off() - """ % (DicoPath['Rdendro'], DicoPath['dendro1'], DicoPath['arbre1']) + """ % (ffr(DicoPath['Rdendro']), ffr(DicoPath['dendro1']), ffr(DicoPath['arbre1'])) if classif_mode == 0: txt += """ @@ -212,21 +238,22 @@ def RchdTxt(DicoPath, RscriptPath, mincl, classif_mode, nbt = 9, libsvdc = False plot(tree.cut2$tree.cl) dev.off() open_file_graph("%s", width = 600, height=400) - plot(tree.cut1$dendro_tot_cl) + plot(tree.cut2$dendro_tot_cl) dev.off() - """ % (DicoPath['dendro2'], DicoPath['arbre2']) + """ % (ffr(DicoPath['dendro2']), ffr(DicoPath['arbre2'])) txt += """ - save.image(file="%s") - """ % DicoPath['RData'] + + #save.image(file="%s") + """ % (ffr(DicoPath['RData'])) + fileout = open(DicoPath['Rchdtxt'], 'w') fileout.write(txt) fileout.close() def RPamTxt(corpus, RscriptPath): - DicoPath = corpus.dictpathout - param = corpus.parametre - print param + DicoPath = corpus.pathout + param = corpus.parametres txt = """ source("%s") """ % (RscriptPath['pamtxt']) @@ -256,7 +283,7 @@ def RchdQuest(DicoPath, RscriptPath, nbcl = 10, mincl = 10): source("%s") source("%s") source("%s") - """ % (RscriptPath['CHD'], RscriptPath['chdquest'], RscriptPath['anacor'],RscriptPath['Rgraph']) + """ % (ffr(RscriptPath['CHD']), ffr(RscriptPath['chdquest']), ffr(RscriptPath['anacor']),ffr(RscriptPath['Rgraph'])) txt += """ nbt <- %i - 1 @@ -267,14 +294,14 @@ def RchdQuest(DicoPath, RscriptPath, nbcl = 10, mincl = 10): chd.result<-Rchdquest("%s","%s","%s", nbt = nbt, mincl = mincl) n1 <- chd.result$n1 classeuce1 <- chd.result$cuce1 - """ % (DicoPath['Act01'], DicoPath['listeuce1'], DicoPath['uce']) + """ % (ffr(DicoPath['mat01.csv']), ffr(DicoPath['listeuce1']), ffr(DicoPath['uce'])) txt += """ tree_tot1 <- make_tree_tot(chd.result$chd) open_file_graph("%s", width = 600, height=400) plot(tree_tot1$tree.cl) dev.off() - """%DicoPath['arbre1'] + """ % ffr(DicoPath['arbre1']) txt += """ tree_cut1 <- make_dendro_cut_tuple(tree_tot1$dendro_tuple, chd.result$coord_ok, classeuce1, 1, nbt) @@ -282,27 +309,28 @@ def RchdQuest(DicoPath, RscriptPath, nbcl = 10, mincl = 10): save(tree.cut1, file="%s") open_file_graph("%s", width = 600, height=400) classes<-n1[,ncol(n1)] - plot.dendropr(tree_cut1$tree.cl,classes) - """ % (DicoPath['Rdendro'],DicoPath['dendro1']) + plot.dendropr(tree_cut1$tree.cl,classes, histo = TRUE) + """ % (ffr(DicoPath['Rdendro']), ffr(DicoPath['dendro1'])) txt += """ save.image(file="%s") - """ % DicoPath['RData'] + """ % ffr(DicoPath['RData']) fileout = open(DicoPath['Rchdquest'], 'w') fileout.write(txt) fileout.close() -def AlcesteTxtProf(DictChdTxtOut, RscriptsPath, clnb, taillecar): +def ReinertTxtProf(DictChdTxtOut, RscriptsPath, clnb, taillecar): txt = "clnb<-%i\n" % clnb txt += """ source("%s") -load("%s") -""" % (RscriptsPath['chdfunct'], DictChdTxtOut['RData']) +#load("%s") +n1 <- read.csv2("%s") +""" % (ffr(RscriptsPath['chdfunct']), ffr(DictChdTxtOut['RData']), ffr(DictChdTxtOut['n1.csv'])) txt += """ dataact<-read.csv2("%s", header = FALSE, sep = ';',quote = '\"', row.names = 1, na.strings = 'NA') datasup<-read.csv2("%s", header = FALSE, sep = ';',quote = '\"', row.names = 1, na.strings = 'NA') dataet<-read.csv2("%s", header = FALSE, sep = ';',quote = '\"', row.names = 1, na.strings = 'NA') -""" % (DictChdTxtOut['Contout'], DictChdTxtOut['ContSupOut'], DictChdTxtOut['ContEtOut']) +""" % (ffr(DictChdTxtOut['Contout']), ffr(DictChdTxtOut['ContSupOut']), ffr(DictChdTxtOut['ContEtOut'])) txt += """ tablesqrpact<-BuildProf(as.matrix(dataact),n1,clnb) tablesqrpsup<-BuildProf(as.matrix(datasup),n1,clnb) @@ -310,7 +338,7 @@ tablesqrpet<-BuildProf(as.matrix(dataet),n1,clnb) """ txt += """ PrintProfile(n1,tablesqrpact[4],tablesqrpet[4],tablesqrpact[5],tablesqrpet[5],clnb,"%s","%s",tablesqrpsup[4],tablesqrpsup[5]) -""" % (DictChdTxtOut['PROFILE_OUT'], DictChdTxtOut['ANTIPRO_OUT']) +""" % (ffr(DictChdTxtOut['PROFILE_OUT']), ffr(DictChdTxtOut['ANTIPRO_OUT'])) txt += """ colnames(tablesqrpact[[2]])<-paste('classe',1:clnb,sep=' ') colnames(tablesqrpact[[1]])<-paste('classe',1:clnb,sep=' ') @@ -327,7 +355,7 @@ write.csv2(chistabletot,file="%s") write.csv2(ptabletot,file="%s") gbcluster<-n1 write.csv2(gbcluster,file="%s") -""" % (DictChdTxtOut['chisqtable'], DictChdTxtOut['ptable'], DictChdTxtOut['SbyClasseOut']) +""" % (ffr(DictChdTxtOut['chisqtable']), ffr(DictChdTxtOut['ptable']), ffr(DictChdTxtOut['SbyClasseOut'])) if clnb > 2 : txt += """ library(ca) @@ -347,7 +375,7 @@ write.csv2(gbcluster,file="%s") #FIXME : split this!!! txt += """ source("%s") - """ % RscriptsPath['Rgraph'] + """ % ffr(RscriptsPath['Rgraph']) txt += """ afc <- summary.ca.dm(afc) @@ -355,36 +383,31 @@ write.csv2(gbcluster,file="%s") write.csv2(afc_table$facteur, file = "%s") write.csv2(afc_table$colonne, file = "%s") write.csv2(afc_table$ligne, file = "%s") - """ % (DictChdTxtOut['afc_facteur'], DictChdTxtOut['afc_col'], DictChdTxtOut['afc_row']) - - txt += """ - xlab <- paste('facteur 1 - ', round(afc$facteur[1,2],2), sep = '') - ylab <- paste('facteur 2 - ', round(afc$facteur[2,2],2), sep = '') - xlab <- paste(xlab, ' %', sep = '') - ylab <- paste(ylab, ' %', sep = '') - """ + """ % (ffr(DictChdTxtOut['afc_facteur']), ffr(DictChdTxtOut['afc_col']), ffr(DictChdTxtOut['afc_row'])) txt += """ PARCEX<-%s """ % taillecar txt += """ - PlotAfc2dCoul(afc, as.data.frame(chistabletot), "%s", what='coord', deb=1, fin=(debsup-1), xlab = xlab, ylab = ylab) - """ % (DictChdTxtOut['AFC2DL_OUT']) + xyminmax <- PlotAfc2dCoul(afc, as.data.frame(chistabletot), "%s", what='coord', deb=1, fin=(debsup-1), xlab = xlab, ylab = ylab) + """ % (ffr(DictChdTxtOut['AFC2DL_OUT'])) txt += """ - PlotAfc2dCoul(afc, as.data.frame(chistabletot), "%s", what='coord', deb=debsup, fin=(debet-1), xlab = xlab, ylab = ylab) - """ % (DictChdTxtOut['AFC2DSL_OUT']) + PlotAfc2dCoul(afc, as.data.frame(chistabletot), "%s", what='coord', deb=debsup, fin=(debet-1), xlab = xlab, ylab = ylab, xmin = xyminmax$xminmax[1], xmax = xyminmax$xminmax[2], ymin = xyminmax$yminmax[1], ymax = xyminmax$yminmax[2], active=FALSE) + """ % (ffr(DictChdTxtOut['AFC2DSL_OUT'])) txt += """ - PlotAfc2dCoul(afc, as.data.frame(chistabletot), "%s", what='coord', deb=debet, fin=fin, xlab = xlab, ylab = ylab) - """ % (DictChdTxtOut['AFC2DEL_OUT']) - txt += """ - PlotAfc2dCoul(afc, as.data.frame(chistabletot), "%s", col=TRUE, what='coord', xlab = xlab, ylab = ylab) - """ % (DictChdTxtOut['AFC2DCL_OUT']) + if ((fin - debet) > 2) { + PlotAfc2dCoul(afc, as.data.frame(chistabletot), "%s", what='coord', deb=debet, fin=fin, xlab = xlab, ylab = ylab, xmin = xyminmax$xminmax[1], xmax = xyminmax$xminmax[2], ymin = xyminmax$yminmax[1], ymax = xyminmax$yminmax[2], active = FALSE) + } + """ % (ffr(DictChdTxtOut['AFC2DEL_OUT'])) txt += """ - PlotAfc2dCoul(afc, as.data.frame(chistabletot), "%s", what='crl', deb=1, fin=(debsup-1), xlab = xlab, ylab = ylab) - PlotAfc2dCoul(afc, as.data.frame(chistabletot), "%s", what='crl', deb=debsup, fin=(debet-1), xlab = xlab, ylab = ylab) - PlotAfc2dCoul(afc, as.data.frame(chistabletot), "%s", what='crl', deb=debet, fin=fin, xlab = xlab, ylab = ylab) - PlotAfc2dCoul(afc, as.data.frame(chistabletot), "%s", col=TRUE, what='crl', xlab = xlab, ylab = ylab) - """ % (DictChdTxtOut['AFC2DCoul'], DictChdTxtOut['AFC2DCoulSup'], DictChdTxtOut['AFC2DCoulEt'], DictChdTxtOut['AFC2DCoulCl']) + PlotAfc2dCoul(afc, as.data.frame(chistabletot), "%s", col=TRUE, what='coord', xlab = xlab, ylab = ylab, xmin = xyminmax$xminmax[1], xmax = xyminmax$xminmax[2], ymin = xyminmax$yminmax[1], ymax = xyminmax$yminmax[2], active=FALSE) + """ % (ffr(DictChdTxtOut['AFC2DCL_OUT'])) +# txt += """ + # PlotAfc2dCoul(afc, as.data.frame(chistabletot), "%s", what='crl', deb=1, fin=(debsup-1), xlab = xlab, ylab = ylab) + # PlotAfc2dCoul(afc, as.data.frame(chistabletot), "%s", what='crl', deb=debsup, fin=(debet-1), xlab = xlab, ylab = ylab) + # PlotAfc2dCoul(afc, as.data.frame(chistabletot), "%s", what='crl', deb=debet, fin=fin, xlab = xlab, ylab = ylab) + # PlotAfc2dCoul(afc, as.data.frame(chistabletot), "%s", col=TRUE, what='crl', xlab = xlab, ylab = ylab) + # """ % (DictChdTxtOut['AFC2DCoul'], DictChdTxtOut['AFC2DCoulSup'], DictChdTxtOut['AFC2DCoulEt'], DictChdTxtOut['AFC2DCoulCl']) txt += """ #rm(dataact) @@ -394,7 +417,7 @@ rm(tablesqrpact) rm(tablesqrpsup) rm(tablesqrpet) save.image(file="%s") -""" % DictChdTxtOut['RData'] +""" % ffr(DictChdTxtOut['RData']) file = open(DictChdTxtOut['RTxtProfGraph'], 'w') file.write(txt) file.close() @@ -410,17 +433,23 @@ def write_afc_graph(self): if self.param['do_select_chi'] : do_select_chi = 'TRUE' else : do_select_chi = 'FALSE' + if self.param['do_select_chi_classe'] : do_select_chi_classe = 'TRUE' + else : do_select_chi_classe = 'FALSE' + if self.param['cex_txt'] : cex_txt = 'TRUE' else : cex_txt = 'FALSE' if self.param['tchi'] : tchi = 'TRUE' else : tchi = 'FALSE' + if self.param['svg'] : svg = 'TRUE' + else : svg = 'FALSE' + with open(self.RscriptsPath['afc_graph'], 'r') as f: txt = f.read() # self.DictPathOut['RData'], \ - scripts = txt % (self.RscriptsPath['Rgraph'],\ + scripts = txt % (ffr(self.RscriptsPath['Rgraph']),\ self.param['typegraph'], \ self.param['what'], \ self.param['facteur'][0],\ @@ -431,6 +460,8 @@ def write_afc_graph(self): self.param['select_nb'], \ do_select_chi, \ self.param['select_chi'], \ + do_select_chi_classe, \ + self.param['nbchic'], \ cex_txt, \ self.param['txt_min'], \ self.param['txt_max'], \ @@ -443,7 +474,8 @@ def write_afc_graph(self): tchi,\ self.param['tchi_min'],\ self.param['tchi_max'],\ - ffr(os.path.dirname(self.fileout))) + ffr(os.path.dirname(self.fileout)),\ + svg) return scripts def print_simi3d(self): @@ -453,10 +485,12 @@ def print_simi3d(self): movie = "'" + ffr(os.path.dirname(self.DictPathOut['RData'])) + "'" else : movie = 'NULL' - if self.section == 'chd_dist_quest' : - header = 'TRUE' - else : - header = 'FALSE' + + #if self.corpus.parametres['type'] == 'corpus' : + # header = 'TRUE' + #else : + # header = 'FALSE' + header = 'FALSE' txt += """ dm<-read.csv2("%s",row.names=1,header = %s) load("%s") @@ -503,66 +537,110 @@ def dendroandbarplot(table, rownames, colnames, rgraph, tmpgraph, intxt = False, """ % (ffr(dendro),ffr(rgraph), ffr(tmpgraph)) return txt -def barplot(table, rownames, colnames, rgraph, tmpgraph, intxt = False) : +def barplot(table, parametres, intxt = False) : if not intxt : txttable = 'c(' + ','.join([','.join(line) for line in table]) + ')' #width = 100 + (15 * len(rownames)) + (100 * len(colnames)) #height = len(rownames) * 15 - rownb = len(rownames) + rownb = len(parametres['rownames']) #if height < 400 : # height = 400 - rownames = 'c("' + '","'.join(rownames) + '")' - colnames = 'c("' + '","'.join(colnames) + '")' + rownames = 'c("' + '","'.join(parametres['rownames']) + '")' + colnames = 'c("' + '","'.join(parametres['colnames']) + '")' + if not intxt : #FIXME txt = """ - inf <- NA di <- matrix(data=%s, nrow=%i, byrow = TRUE) - di[is.na(di)] <- max(di, na.rm=TRUE) + 2 + toinf <- which(di == Inf) + tominf <- which(di == -Inf) + if (length(toinf)) { + di[toinf] <- NA + valmax <- max(di, na.rm = TRUE) + if (valmax <= 0) { + valmax <- 2 + } else { + valmax <- valmax + 2 + } + di[toinf] <- valmax + } + if (length(tominf)) { + di[tominf] <- NA + valmin <- min(di, na.rm = TRUE) + if (valmin >=0) { + valmin <- -2 + } else { + valmin <- valmin - 2 + } + di[tominf] <- valmin + } rownames(di)<- %s colnames(di) <- %s """ % (txttable, rownb, rownames, colnames) else : txt = intxt - txt += """ - source("%s") - color = rainbow(nrow(di)) - width <- 100 + (20*length(rownames(di))) + (100 * length(colnames(di))) - height <- nrow(di) * 15 - if (height < 400) { height <- 400} - open_file_graph("%s",width = width, height = height) - par(mar=c(0,0,0,0)) - layout(matrix(c(1,2),1,2, byrow=TRUE),widths=c(3,lcm(7))) - par(mar=c(2,2,1,0)) - coord <- barplot(as.matrix(di), beside = TRUE, col = color, space = c(0.1,0.6)) - c <- colMeans(coord) - c1 <- c[-1] - c2 <- c[-length(c)] - cc <- cbind(c1,c2) - lcoord <- apply(cc, 1, mean) - abline(v=lcoord) - if (min(di) < 0) { - amp <- abs(max(di) - min(di)) - } else { - amp <- max(di) - } - if (amp < 10) { - d <- 2 - } else { - d <- signif(amp%%/%%10,1) - } - mn <- round(min(di)) - mx <- round(max(di)) - for (i in mn:mx) { - if ((i/d) == (i%%/%%d)) { - abline(h=i,lty=3) + if not 'tree' in parametres : + txt += """ + source("%s") + color = rainbow(nrow(di)) + width <- %i + height <- %i + open_file_graph("%s",width = width, height = height, svg = %s) + par(mar=c(0,0,0,0)) + layout(matrix(c(1,2),1,2, byrow=TRUE),widths=c(3,lcm(7))) + par(mar=c(8,4,1,0)) + yp = ifelse(length(toinf), 0.2, 0) + ym = ifelse(length(tominf), 0.2, 0) + ymin <- ifelse(!length(which(di < 0)), 0, min(di) - ym) + coord <- barplot(as.matrix(di), beside = TRUE, col = color, space = c(0.1,0.6), ylim=c(ymin, max(di) + yp), las = 2) + if (length(toinf)) { + coordinf <- coord[toinf] + valinf <- di[toinf] + text(x=coordinf, y=valinf + 0.1, 'i') } - } - par(mar=c(0,0,0,0)) - plot(0, axes = FALSE, pch = '') - legend(x = 'center' , rownames(di), fill = color) - dev.off() - """ % (rgraph, ffr(tmpgraph)) + if (length(tominf)) { + coordinf <- coord[toinf] + valinf <- di[toinf] + text(x=coordinf, y=valinf - 0.1, 'i') + } + c <- colMeans(coord) + c1 <- c[-1] + c2 <- c[-length(c)] + cc <- cbind(c1,c2) + lcoord <- apply(cc, 1, mean) + abline(v=lcoord) + if (min(di) < 0) { + amp <- abs(max(di) - min(di)) + } else { + amp <- max(di) + } + if (amp < 10) { + d <- 2 + } else { + d <- signif(amp%%/%%10,1) + } + mn <- round(min(di)) + mx <- round(max(di)) + for (i in mn:mx) { + if ((i/d) == (i%%/%%d)) { + abline(h=i,lty=3) + } + } + par(mar=c(0,0,0,0)) + plot(0, axes = FALSE, pch = '') + legend(x = 'center' , rownames(di), fill = color) + dev.off() + """ % (ffr(parametres['rgraph']), parametres['width'], parametres['height'], ffr(parametres['tmpgraph']), parametres['svg']) + else : + txt += """ + load("%s") + library(ape) + source("%s") + width = %i + height = %i + open_file_graph("%s", width=width, height=height, svg = %s) + plot.dendro.lex(tree.cut1$tree.cl, di) + """ % (ffr(parametres['tree']), ffr(parametres['rgraph']), parametres['width'], parametres['height'], ffr(parametres['tmpgraph']), parametres['svg']) return txt #def RAfcUci(DictAfcUciOut, nd=2, RscriptsPath='', PARCEX='0.8'): @@ -620,3 +698,565 @@ def barplot(table, rownames, colnames, rgraph, tmpgraph, intxt = False) : # f.write(txt) # f.close() +class PrintSimiScript(PrintRScript) : + def make_script(self) : + self.txtgraph = '' + self.packages(['igraph', 'proxy', 'Matrix']) + self.sources([self.analyse.parent.RscriptsPath['simi'], self.analyse.parent.RscriptsPath['Rgraph']]) + txt = '' + if not self.parametres['keep_coord'] and not (self.parametres['type'] == 'simimatrix' or self.parametres['type'] == 'simiclustermatrix') : + txt += """ + dm.path <- "%s" + cn.path <- "%s" + selected.col <- "%s" + """ % (ffr(self.pathout['mat01.csv']), ffr(self.pathout['actives.csv']), ffr(self.pathout['selected.csv'])) + if 'word' in self.parametres : + txt += """ + word <- TRUE + index <- %i + 1 + """ % self.parametres['word'] + else : + txt += """ + word <- FALSE + """ + txt += """ + dm <-readMM(dm.path) + cn <- read.table(cn.path, sep='\t', quote='"') + colnames(dm) <- cn[,1] + if (file.exists(selected.col)) { + sel.col <- read.csv2(selected.col, header = FALSE) + sel.col <- sel.col[,1] + 1 + } else { + sel.col <- 1:ncol(dm) + } + if (!word) { + dm <- dm[, sel.col] + } else { + forme <- colnames(dm)[index] + if (!index %in% sel.col) { + sel.col <- append(sel.col, index) + } + dm <- dm[, sel.col] + index <- which(colnames(dm) == forme) + } + """ + elif not self.parametres['keep_coord'] and (self.parametres['type'] == 'simimatrix' or self.parametres['type'] == 'simiclustermatrix'): + txt += """ + dm.path <- "%s" + selected.col <- "%s" + """ % (ffr(self.pathout['mat01.csv']), ffr(self.pathout['selected.csv'])) + if 'word' in self.parametres : + txt += """ + word <- TRUE + index <- %i + 1 + """ % self.parametres['word'] + else : + txt += """ + word <- FALSE + """ + txt += """ + dm <-read.csv2(dm.path) + dm <- as.matrix(dm) + if (file.exists(selected.col)) { + sel.col <- read.csv2(selected.col, header = FALSE) + sel.col <- sel.col[,1] + 1 + } else { + sel.col <- 1:ncol(dm) + } + if (!word) { + dm <- dm[, sel.col] + } else { + forme <- colnames(dm)[index] + if (!index %in% sel.col) { + sel.col <- append(sel.col, index) + } + dm <- dm[, sel.col] + index <- which(colnames(dm) == forme) + } + """ + else : + txt += """ + load("%s") + """ % ffr(self.pathout['RData.RData']) + + if self.parametres['coeff'] == 0 : + method = 'cooc' + if not self.parametres['keep_coord'] : + txt += """ + method <- 'cooc' + mat <- make.a(dm) + """ + else : + if not self.parametres['keep_coord'] : + txt += """ + dm <- as.matrix(dm) + """ + if self.parametres['coeff'] == 1 : + method = 'prcooc' + txt += """ + method <- 'Russel' + mat <- simil(dm, method = 'Russel', diag = TRUE, upper = TRUE, by_rows = FALSE) + """ + elif self.analyse.indices[self.parametres['coeff']] == 'binomial' : + method = 'binomial' + if not self.parametres['keep_coord'] : + txt += """ + method <- 'binomial' + mat <- binom.sim(dm) + """ + elif self.parametres['coeff'] != 0 : + method = self.analyse.indices[self.parametres['coeff']] + if not self.parametres['keep_coord'] : + txt += """ + method <-"%s" + mat <- simil(dm, method = method, diag = TRUE, upper = TRUE, by_rows = FALSE) + """ % self.analyse.indices[self.parametres['coeff']] + if not self.parametres['keep_coord'] : + txt += """ + mat <- as.matrix(stats::as.dist(mat,diag=TRUE,upper=TRUE)) + mat[is.na(mat)] <- 0 + if (length(which(mat == Inf))) { + infp <- which(mat == Inf) + mat[infp] <- NA + maxmat <- max(mat, na.rm = TRUE) + if (maxmat > 0) { + maxmat <- maxmat + 1 + } else { + maxmat <- 0 + } + mat[infp] <- maxmat + } + if (length(which(mat == -Inf))) { + infm <- which(mat == -Inf) + mat[infm] <- NA + minmat <- min(mat, na.rm = TRUE) + if (maxmat < 0) { + minmat <- minmat - 1 + } else { + minmat <- 0 + } + mat[infm] <- minmat + } + """ + if 'word' in self.parametres and not self.parametres['keep_coord'] : + txt += """ + mat <- graph.word(mat, index) + cs <- colSums(mat) + if (length(cs)) mat <- mat[,-which(cs==0)] + rs <- rowSums(mat) + if (length(rs)) mat <- mat[-which(rs==0),] + if (length(cs)) dm <- dm[, -which(cs==0)] + """ + + if self.parametres['layout'] == 0 : layout = 'random' + if self.parametres['layout'] == 1 : layout = 'circle' + if self.parametres['layout'] == 2 : layout = 'frutch' + if self.parametres['layout'] == 3 : layout = 'kawa' + if self.parametres['layout'] == 4 : layout = 'graphopt' + + + self.filename='' + if self.parametres['type_graph'] == 0 : type = 'tkplot' + if self.parametres['type_graph'] == 1 : + graphnb = 1 + type = 'nplot' + dirout = os.path.dirname(self.pathout['mat01.csv']) + while os.path.exists(os.path.join(dirout,'graph_simi_'+str(graphnb)+'.png')): + graphnb +=1 + self.filename = ffr(os.path.join(dirout,'graph_simi_'+str(graphnb)+'.png')) + if self.parametres['type_graph'] == 2 : type = 'rgl' + if self.parametres['type_graph'] == 3 : + graphnb = 1 + type = 'web' + dirout = os.path.dirname(self.pathout['mat01.csv']) + while os.path.exists(os.path.join(dirout,'web_'+str(graphnb))): + graphnb +=1 + self.filename = ffr(os.path.join(dirout,'web_'+str(graphnb))) + os.mkdir(self.filename) + self.filename = os.path.join(self.filename, 'gexf.gexf') + if self.parametres['type_graph'] == 4 : + graphnb = 1 + type = 'rglweb' + dirout = os.path.dirname(self.pathout['mat01.csv']) + while os.path.exists(os.path.join(dirout,'webrgl_'+str(graphnb))): + graphnb +=1 + self.filename = ffr(os.path.join(dirout,'webrgl_'+str(graphnb))) + os.mkdir(self.filename) + + if self.parametres['arbremax'] : + arbremax = 'TRUE' + self.txtgraph += ' - arbre maximum' + else : arbremax = 'FALSE' + + if self.parametres['coeff_tv'] : + coeff_tv = self.parametres['coeff_tv_nb'] + tvminmax = 'c(NULL,NULL)' + elif not self.parametres['coeff_tv'] or self.parametres.get('sformchi', False) : + coeff_tv = 'NULL' + tvminmax = 'c(%i, %i)' %(self.parametres['tvmin'], self.parametres['tvmax']) + if self.parametres['coeff_te'] : coeff_te = 'c(%i,%i)' % (self.parametres['coeff_temin'], self.parametres['coeff_temax']) + else : coeff_te = 'NULL' + + if self.parametres['vcex'] or self.parametres.get('cexfromchi', False) : + vcexminmax = 'c(%i/10,%i/10)' % (self.parametres['vcexmin'],self.parametres['vcexmax']) + else : + vcexminmax = 'c(NULL,NULL)' + if not self.parametres['label_v'] : label_v = 'FALSE' + else : label_v = 'TRUE' + + if not self.parametres['label_e'] : label_e = 'FALSE' + else : label_e = 'TRUE' + + if self.parametres['seuil_ok'] : seuil = str(self.parametres['seuil']) + else : seuil = 'NULL' + + if not self.parametres.get('edgecurved', False) : + ec = 'FALSE' + else : + ec = 'TRUE' + + txt += """ + edge.curved <- %s + """ % ec + + cols = str(self.parametres['cols']).replace(')',', max=255)') + cola = str(self.parametres['cola']).replace(')',',max=255)') + + txt += """ + minmaxeff <- %s + """ % tvminmax + txt += """ + vcexminmax <- %s + """ % vcexminmax + txt += """ + cex = %i/10 + """ % self.parametres['cex'] + + if self.parametres['film'] : + txt += """ + film <- "%s" + """ % self.pathout['film'] + else : + txt += """ + film <- NULL + """ + txt += """ + seuil <- %s + if (!is.null(seuil)) { + if (method!='cooc') { + seuil <- seuil/100 + } + } + """ % seuil + + txt += """ + label.v <- %s + label.e <- %s + """ % (label_v, label_e) + txt += """ + cols <- rgb%s + cola <- rgb%s + """ % (cols, cola) + txt += """ + width <- %i + height <- %i + """ % (self.parametres['width'], self.parametres['height']) + if self.parametres['keep_coord'] : + txt += """ + coords <- try(coords, TRUE) + if (!is.matrix(coords)) { + coords<-NULL + } + """ + else : + txt += """ + coords <- NULL + """ + txt += """ + alpha <- %i/100 + """ % self.parametres['alpha'] + txt += """ + alpha <- %i/100 + """ % self.parametres['alpha'] +############################################# + if self.parametres.get('bystar',False) : + txt += """ + et <- list() + """ + for i, line in enumerate(self.parametres['listet']) : + txt+= """ + et[[%i]] <- c(%s) + """ % (i+1, ','.join([`val + 1` for val in line])) + txt+= """ + unetoile <- c('%s') + """ % ("','".join([val for val in self.parametres['selectedstars']])) + txt += """ + fsum <- NULL + rs <- rowSums(dm) + for (i in 1:length(unetoile)) { + print(unetoile[i]) + tosum <- et[[i]] + if (length(tosum) > 1) { + fsum <- cbind(fsum, colSums(dm[tosum,])) + } else { + fsum <- cbind(fsum, dm[tosum,]) + } + } + source("%s") + lex <- AsLexico2(fsum, chip=TRUE) + dcol <- apply(lex[[4]],1,which.max) + toblack <- apply(lex[[4]],1,max) + gcol <- rainbow(length(unetoile)) + #gcol[2] <- 'orange' + vertex.label.color <- gcol[dcol] + vertex.label.color[which(toblack <= 3.84)] <- 'black' + leg <- list(unetoile=unetoile, gcol=gcol) + cols <- vertex.label.color + chivertex.size <- norm.vec(toblack, vcexminmax[1], vcexminmax[2]) + + """ % (ffr(self.analyse.parent.RscriptsPath['chdfunct'])) + else : + txt += """ + vertex.label.color <- 'black' + chivertex.size <- 1 + leg<-NULL + """ +############################################# + +# txt += """ +# eff <- colSums(dm) +# g.ori <- graph.adjacency(mat, mode='lower', weighted = TRUE) +# w.ori <- E(g.ori)$weight +# if (max.tree) { +# if (method == 'cooc') { +# E(g.ori)$weight <- 1 / w.ori +# } else { +# E(g.ori)$weigth <- 1 - w.ori +# } +# g.max <- minimum.spanning.tree(g.ori) +# if (method == 'cooc') { +# E(g.max)$weight <- 1 / E(g.max)$weight +# } else { +# E(g.max)$weight <- 1 - E(g.max)$weight +# } +# g.toplot <- g.max +# } else { +# g.toplot <- g.ori +# } +# """ + if self.parametres['com'] : + com = `self.parametres['communities']` + else : + com = 'NULL' + if self.parametres['halo'] : + halo = 'TRUE' + else : + halo = 'FALSE' + txt += """ + communities <- %s + halo <- %s + """ % (com, halo) + txt += """ + eff <- colSums(dm) + x <- list(mat = mat, eff = eff) + graph.simi <- do.simi(x, method='%s', seuil = seuil, p.type = '%s', layout.type = '%s', max.tree = %s, coeff.vertex=%s, coeff.edge = %s, minmaxeff = minmaxeff, vcexminmax = vcexminmax, cex = cex, coords = coords, communities = communities, halo = halo) + """ % (method, type, layout, arbremax, coeff_tv, coeff_te) + + if self.parametres.get('bystar',False) : + if self.parametres.get('cexfromchi', False) : + txt+=""" + label.cex<-chivertex.size + """ + else : + txt+=""" + label.cex <- cex + """ + if self.parametres.get('sfromchi', False) : + txt += """ + vertex.size <- norm.vec(toblack, minmaxeff[1], minmaxeff[2]) + """ + else : + txt += """ + vertex.size <- NULL + """ + else : + #print self.parametres + if (self.parametres['type'] == 'clustersimitxt' and self.parametres.get('tmpchi', False)) or (self.parametres['type'] in ['simimatrix','simiclustermatrix'] and 'tmpchi' in self.parametres): + txt += """ + lchi <- read.table("%s") + lchi <- lchi[,1] + """ % ffr(self.parametres['tmpchi']) + txt += """ + lchi <- lchi[sel.col] + """ + if self.parametres['type'] in ['clustersimitxt', 'simimatrix', 'simiclustermatrix'] and self.parametres.get('cexfromchi', False) : + txt += """ + label.cex <- norm.vec(lchi, vcexminmax[1], vcexminmax[2]) + """ + else : + txt += """ + if (is.null(vcexminmax[1])) { + label.cex <- cex + } else { + label.cex <- graph.simi$label.cex + } + """ + if (self.parametres['type'] in ['clustersimitxt', 'simimatrix', 'simiclustermatrix']) and self.parametres.get('sfromchi', False): + txt += """ + vertex.size <- norm.vec(lchi, minmaxeff[1], minmaxeff[2]) + if (!length(vertex.size)) vertex.size <- 0 + """ + else : + txt += """ + if (is.null(minmaxeff[1])) { + vertex.size <- 0 + } else { + vertex.size <- graph.simi$eff + } + """ + #txt += """ vertex.size <- NULL """ + if self.parametres['svg'] : svg = 'TRUE' + else : svg = 'FALSE' + txt += """ + svg <- %s + """ % svg + txt += """ + vertex.col <- cols + if (!is.null(graph.simi$com)) { + com <- graph.simi$com + colm <- rainbow(length(com)) + if (vertex.size != 0 || graph.simi$halo) { + vertex.label.color <- 'black' + vertex.col <- colm[membership(com)] + } else { + vertex.label.color <- colm[membership(com)] + } + } + coords <- plot.simi(graph.simi, p.type='%s',filename="%s", vertex.label = label.v, edge.label = label.e, vertex.col = vertex.col, vertex.label.color = vertex.label.color, vertex.label.cex=label.cex, vertex.size = vertex.size, edge.col = cola, leg=leg, width = width, height = height, alpha = alpha, movie = film, edge.curved = edge.curved, svg = svg) + save.image(file="%s") + """ % (type, self.filename, ffr(self.pathout['RData'])) + + self.add(txt) + self.write() + +class WordCloudRScript(PrintRScript) : + def make_script(self) : + self.sources([self.analyse.parent.RscriptsPath['Rgraph']]) + self.packages(['wordcloud']) + bg_col = Rcolor(self.parametres['col_bg']) + txt_col = Rcolor(self.parametres['col_text']) + if self.parametres['svg'] : + svg = 'TRUE' + else : + svg = 'FALSE' + txt = """ + svg <- %s + """ % svg + txt += """ + act <- read.csv2("%s", header = FALSE, row.names=1, sep='\t') + selected.col <- read.table("%s") + toprint <- as.matrix(act[selected.col[,1] + 1,]) + rownames(toprint) <- rownames(act)[selected.col[,1] + 1] + maxword <- %i + if (nrow(toprint) > maxword) { + toprint <- as.matrix(toprint[order(toprint[,1], decreasing=TRUE),]) + toprint <- as.matrix(toprint[1:maxword,]) + } + open_file_graph("%s", width = %i, height = %i , svg = svg) + par(bg=rgb%s) + wordcloud(row.names(toprint), toprint[,1], scale=c(%f,%f), random.order=FALSE, colors=rgb%s) + dev.off() + """ % (ffr(self.analyse.pathout['actives_eff.csv']), ffr(self.analyse.pathout['selected.csv']), self.parametres['maxword'], ffr(self.parametres['graphout']), self.parametres['width'], self.parametres['height'], bg_col, self.parametres['maxcex'], self.parametres['mincex'], txt_col) + self.add(txt) + self.write() + +class ProtoScript(PrintRScript) : + def make_script(self) : + self.sources([self.analyse.parent.RscriptsPath['Rgraph'], self.analyse.parent.RscriptsPath['prototypical.R']]) + self.packages(['wordcloud']) + if self.parametres.get('cloud', False) : + cloud = 'TRUE' + else : + cloud = 'FALSE' + txt = """ + errorn <- function(x) { + qnorm(0.975)*sd(x)/sqrt(lenght(n)) + } + errort <- function(x) { + qt(0.975,df=lenght(x)-1)*sd(x)/sqrt(lenght(x)) + } + mat <- read.csv2("%s", header = FALSE, row.names=1, sep='\t', quote='"', dec='.') + open_file_graph("%s",height=800, width=1000) + prototypical(mat, mfreq = %s, mrank = %s, cloud = FALSE, cexrange=c(1,2.4), cexalpha= c(0.4, 1), type = '%s') + dev.off() + """ % (self.analyse.pathout['table.csv'], self.analyse.pathout['proto.png'], self.parametres['limfreq'], self.parametres['limrang'], self.parametres['typegraph']) + self.add(txt) + self.write() + + +class ExportAfc(PrintRScript) : + def make_script(self) : + self.source([self.analyse.parent.RscriptsPath['Rgraph']]) + self.packages(['rgexf']) + txt = """ + """ + +class TgenSpecScript(PrintRScript): + def make_script(self): + self.packages(['textometry']) + txt = """ + tgen <- read.csv2("%s", row.names = 1, sep = '\\t') + """ % ffr(self.parametres['tgeneff']) + txt += """ + tot <- tgen[nrow(tgen), ] + result <- NULL + tgen <- tgen[-nrow(tgen),] + for (i in 1:nrow(tgen)) { + mat <- rbind(tgen[i,], tot - tgen[i,]) + specmat <- specificities(mat) + result <- rbind(result, specmat[1,]) + } + colnames(result) <- colnames(tgen) + row.names(result) <- rownames(tgen) + write.table(result, file = "%s", sep='\\t', col.names = NA) + """ % ffr(self.pathout['tgenspec.csv']) + self.add(txt) + +class TgenProfScript(PrintRScript): + def make_script(self): + self.sources([self.analyse.ira.RscriptsPath['chdfunct']]) + txt = """ + tgen <- read.csv2("%s", row.names = 1, sep = '\\t') + """ % ffr(self.parametres['tgeneff']) + txt += """ + res <- build.prof.tgen(tgen) + write.table(res$chi2, file = "%s", sep='\\t', col.names = NA) + write.table(res$pchi2, file = "%s", sep='\\t', col.names = NA) + """ % (ffr(self.pathout['tgenchi2.csv']), ffr(self.pathout['tgenpchi2.csv'])) + self.add(txt) + +class FreqMultiScript(PrintRScript): + def make_script(self): + self.sources([self.analyse.parent.RscriptsPath['Rgraph']]) + txt = """ + freq <- read.csv2("%s", row.names=1, sep='\\t', dec='.') + """ % ffr(self.pathout['frequences.csv']) + txt += """ + toplot <- freq[order(freq[,2]) ,2] + toplot.names = rownames(freq)[order(freq[,2])] + h <- 80 + (20 * nrow(freq)) + open_file_graph("%s",height=h, width=500) + par(mar=c(3,20,3,3)) + barplot(toplot, names = toplot.names, horiz=TRUE, las =1, col = rainbow(nrow(freq))) + dev.off() + """ % ffr(self.pathout['barplotfreq.png']) + txt += """ + toplot <- freq[order(freq[,4]) ,4] + toplot.names = rownames(freq)[order(freq[,4])] + open_file_graph("%s",height=h, width=500) + par(mar=c(3,20,3,3)) + barplot(toplot, names = toplot.names, horiz=TRUE, las =1, col = rainbow(nrow(freq))) + dev.off() + """ % ffr(self.pathout['barplotrow.png']) + self.add(txt) + self.write() \ No newline at end of file