X-Git-Url: http://iramuteq.org/git?p=iramuteq;a=blobdiff_plain;f=PrintRScript.py;h=16cf7defde7bcc19e7e417f4c90ef4a4087c9ade;hp=8e54bf42bcfc315eb0b92648fd8dbd39aa4c02c4;hb=287f9e72c3e3d666b016dff0fa3dc39419adfcc2;hpb=b02f2cba3d8c7dba29a54de8b589626d008d2c5b diff --git a/PrintRScript.py b/PrintRScript.py index 8e54bf4..16cf7de 100644 --- a/PrintRScript.py +++ b/PrintRScript.py @@ -4,7 +4,7 @@ #License: GNU/GPL import tempfile -from chemins import ffr +from chemins import ffr, PathOut import os import locale from datetime import datetime @@ -18,12 +18,13 @@ class PrintRScript : self.pathout = analyse.pathout self.analyse = analyse self.parametres = analyse.parametres + #self.scriptout = ffr(self.pathout['lastRscript.R']) self.scriptout = self.pathout['temp'] self.script = u"#Script genere par IRaMuTeQ - %s\n" % datetime.now().ctime() - + def add(self, txt) : self.script = '\n'.join([self.script, txt]) - + def defvar(self, name, value) : self.add(' <- '.join([name, value])) @@ -154,7 +155,7 @@ def RchdTxt(DicoPath, RscriptPath, mincl, classif_mode, nbt = 9, svdmethod = 'sv data1 <- as(data1, "dgCMatrix") row.names(data1) <- 1:nrow(data1) """ % ffr(DicoPath['TableUc1']) - + if classif_mode == 0: txt += """ data2 <- readMM("%s") @@ -162,28 +163,32 @@ def RchdTxt(DicoPath, RscriptPath, mincl, classif_mode, nbt = 9, svdmethod = 'sv row.names(data2) <- 1:nrow(data2) """ % ffr(DicoPath['TableUc2']) txt += """ - chd1<-CHD(data1, x = nbt, mode.patate = mode.patate, svd.method = svd.method, libsvdc.path = libsvdc.path) - """ - + log1 <- "%s" + chd1<-CHD(data1, x = nbt, mode.patate = mode.patate, svd.method = + svd.method, libsvdc.path = libsvdc.path)#, log.file = log1) + """ % ffr(DicoPath['log-chd1.txt']) + if classif_mode == 0: txt += """ - chd2<-CHD(data2, x = nbt, mode.patate = mode.patate, svd.method = svd.method, libsvdc.path = libsvdc.path) - """ - + log2 <- "%s" + chd2<-CHD(data2, x = nbt, mode.patate = mode.patate, svd.method = + svd.method, libsvdc.path = libsvdc.path)#, log.file = log2) + """ % ffr(DicoPath['log-chd2.txt']) + txt += """ #lecture des uce listuce1<-read.csv2("%s") """ % ffr(DicoPath['listeuce1']) - + if classif_mode == 0: txt += """ listuce2<-read.csv2("%s") """ % ffr(DicoPath['listeuce2']) - + txt += """ rm(data1) """ - + if classif_mode == 0: txt += """ rm(data2) @@ -203,14 +208,14 @@ def RchdTxt(DicoPath, RscriptPath, mincl, classif_mode, nbt = 9, svdmethod = 'sv 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$tree.cl) # dev.off() """ % ffr(DicoPath['arbre1']) - + if classif_mode == 0: txt += """ classeuce2 <- chd.result$cuce2 @@ -218,19 +223,19 @@ def RchdTxt(DicoPath, RscriptPath, mincl, classif_mode, nbt = 9, svdmethod = 'sv # open_file_graph("%s", width = 600, height=400) # plot(tree.tot2$tree.cl) # dev.off() - """ % ffr(DicoPath['arbre2'] ) - + """ % ffr(DicoPath['arbre2'] ) + txt += """ tree.cut1 <- make_dendro_cut_tuple(tree.tot1$dendro_tuple, chd.result$coord_ok, classeuce1, 1, nbt) save(tree.cut1, file="%s") - + open_file_graph("%s", width = 600, height=400) plot.dendropr(tree.cut1$tree.cl,classes, histo=TRUE) open_file_graph("%s", width = 600, height=400) plot(tree.cut1$dendro_tot_cl) dev.off() """ % (ffr(DicoPath['Rdendro']), ffr(DicoPath['dendro1']), ffr(DicoPath['arbre1'])) - + if classif_mode == 0: txt += """ tree.cut2 <- make_dendro_cut_tuple(tree.tot2$dendro_tuple, chd.result$coord_ok, classeuce2, 2, nbt) @@ -241,12 +246,12 @@ def RchdTxt(DicoPath, RscriptPath, mincl, classif_mode, nbt = 9, svdmethod = 'sv plot(tree.cut2$dendro_tot_cl) dev.off() """ % (ffr(DicoPath['dendro2']), ffr(DicoPath['arbre2'])) - + txt += """ - + #save.image(file="%s") """ % (ffr(DicoPath['RData'])) - + fileout = open(DicoPath['Rchdtxt'], 'w') fileout.write(txt) fileout.close() @@ -319,7 +324,7 @@ def RchdQuest(DicoPath, RscriptPath, nbcl = 10, mincl = 10): 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") @@ -704,7 +709,7 @@ class PrintSimiScript(PrintRScript) : 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': + if not self.parametres['keep_coord'] and not (self.parametres['type'] == 'simimatrix' or self.parametres['type'] == 'simiclustermatrix') : txt += """ dm.path <- "%s" cn.path <- "%s" @@ -740,7 +745,7 @@ class PrintSimiScript(PrintRScript) : index <- which(colnames(dm) == forme) } """ - elif not self.parametres['keep_coord'] and self.parametres['type'] == 'simimatrix' : + elif not self.parametres['keep_coord'] and (self.parametres['type'] == 'simimatrix' or self.parametres['type'] == 'simiclustermatrix'): txt += """ dm.path <- "%s" selected.col <- "%s" @@ -909,7 +914,16 @@ class PrintSimiScript(PrintRScript) : 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)') @@ -926,7 +940,7 @@ class PrintSimiScript(PrintRScript) : if self.parametres['film'] : txt += """ film <- "%s" - """ % self.pathout['film'] + """ % ffr(self.pathout['film']) else : txt += """ film <- NULL @@ -935,7 +949,7 @@ class PrintSimiScript(PrintRScript) : seuil <- %s if (!is.null(seuil)) { if (method!='cooc') { - seuil <- seuil/100 + seuil <- seuil/1000 } } """ % seuil @@ -1072,7 +1086,7 @@ class PrintSimiScript(PrintRScript) : """ else : #print self.parametres - if (self.parametres['type'] == 'clustersimitxt' and self.parametres.get('tmpchi', False)) or (self.parametres['type'] == 'simimatrix' and 'tmpchi' in 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] @@ -1080,7 +1094,7 @@ class PrintSimiScript(PrintRScript) : txt += """ lchi <- lchi[sel.col] """ - if self.parametres['type'] == 'clustersimitxt' and self.parametres.get('cexfromchi', False) : + if self.parametres['type'] in ['clustersimitxt', 'simimatrix', 'simiclustermatrix'] and self.parametres.get('cexfromchi', False) : txt += """ label.cex <- norm.vec(lchi, vcexminmax[1], vcexminmax[2]) """ @@ -1092,7 +1106,7 @@ class PrintSimiScript(PrintRScript) : label.cex <- graph.simi$label.cex } """ - if (self.parametres['type'] == 'clustersimitxt' or self.parametres['type'] == 'simimatrix') and self.parametres.get('sfromchi', False): + 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 @@ -1123,7 +1137,7 @@ class PrintSimiScript(PrintRScript) : 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, svg = svg) + 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'])) @@ -1180,7 +1194,7 @@ class ProtoScript(PrintRScript) : 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']) + """ % (ffr(self.analyse.pathout['table.csv']), ffr(self.analyse.pathout['proto.png']), self.parametres['limfreq'], self.parametres['limrang'], self.parametres['typegraph']) self.add(txt) self.write() @@ -1192,6 +1206,42 @@ class ExportAfc(PrintRScript) : txt = """ """ +class MergeGraphes(PrintRScript) : + def __init__(self, analyse): + self.script = u"#Script genere par IRaMuTeQ - %s\n" % datetime.now().ctime() + self.pathout = PathOut() + self.parametres = analyse.parametres + self.scriptout = self.pathout['temp'] + self.analyse = analyse + + def make_script(self) : + #FIXME + + txt = """ + library(igraph) + library(Matrix) + graphs <- list() + """ + load = """ + load("%s") + g <- graph.simi$graph + V(g)$weight <- (graph.simi$mat.eff/nrow(dm))*100 + graphs[['%s']] <- g + """ + for i, graph in enumerate(self.parametres['graphs']) : + path = os.path.dirname(graph) + gname = ''.join(['g', `i`]) + RData = os.path.join(path,'RData.RData') + txt += load % (ffr(RData), gname) + self.add(txt) + self.sources([self.analyse.parent.RscriptsPath['simi']]) + txt = """ + ng <- merge.graph(graphs) + ngraph <- list(graph=ng, layout=layout.fruchterman.reingold(ng, dim=3), labex.cex=V(ng)$weight) + write.graph(ng, "%s", format = 'graphml') + """ % ffr(self.parametres['grapheout']) + self.add(txt) + class TgenSpecScript(PrintRScript): def make_script(self): self.packages(['textometry']) @@ -1210,6 +1260,235 @@ class TgenSpecScript(PrintRScript): colnames(result) <- colnames(tgen) row.names(result) <- rownames(tgen) write.table(result, file = "%s", sep='\\t', col.names = NA) - """ % self.pathout['tgenspec.csv'] + """ % ffr(self.pathout['tgenspec.csv']) + self.add(txt) + +class TgenProfScript(PrintRScript): + def make_script(self): + self.sources([self.analyse.ira.RscriptsPath['chdfunct']]) + txt = """ + tgen <- read.csv2("%s", row.names = 1, sep = '\\t') + """ % ffr(self.parametres['tgeneff']) + txt += """ + tgenlem <- read.csv2("%s", row.names = 1, sep = '\\t') + """ % ffr(self.parametres['tgenlemeff']) + txt += """ + res <- build.prof.tgen(tgen) + write.table(res$chi2, file = "%s", sep='\\t', col.names = NA) + write.table(res$pchi2, file = "%s", sep='\\t', col.names = NA) + """ % (ffr(self.pathout['tgenchi2.csv']), ffr(self.pathout['tgenpchi2.csv'])) + txt += """ + reslem <- build.prof.tgen(tgenlem) + write.table(reslem$chi2, file = "%s", sep='\\t', col.names = NA) + write.table(reslem$pchi2, file = "%s", sep='\\t', col.names = NA) + """ % (ffr(self.pathout['tgenlemchi2.csv']), ffr(self.pathout['tgenlempchi2.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() + +class LabbeScript(PrintRScript) : + def make_script(self) : + self.sources([self.analyse.parent.RscriptsPath['distance-labbe.R'], + self.analyse.parent.RscriptsPath['Rgraph']]) + txt = """ + tab <- read.csv2("%s", header=TRUE, sep=';', row.names=1) + """ % (ffr(self.pathout['tableafcm.csv'])) + txt += """ + dist.mat <- dist.labbe(tab) + dist.mat <- as.dist(dist.mat, upper=F, diag=F) + write.table(as.matrix(dist.mat), "%s", sep='\t') + library(cluster) + library(ape) + chd <- hclust(dist.mat, method="ward.D2") + open_file_graph("%s", width=1000, height=1000, svg=F) + par(cex=1.2) + plot.phylo(as.phylo(chd), type='unrooted', lab4ut="axial") + dev.off() + """ % (ffr(self.pathout['distmat.csv']), ffr(self.pathout['labbe-tree.png'])) + txt +=""" + open_file_graph("%s", width=1000, height=1000, svg=F) + par(mar=c(10,1,1,10)) + heatmap(as.matrix(dist.mat), symm = T, distfun=function(x) as.dist(x)) + dev.off() + """ % ffr(self.pathout['labbe-heatmap.png']) + txt += """ + #http://stackoverflow.com/questions/3081066/what-techniques-exists-in-r-to-visualize-a-distance-matrix + dst <- data.matrix(dist.mat) + dim <- ncol(dst) + rn <- row.names(as.matrix(dist.mat)) + open_file_graph("%s", width=1500, height=1000, svg=F) + par(mar=c(10,10,3,3)) + image(1:dim, 1:dim, dst, axes = FALSE, xlab="", ylab="") + axis(1, 1:dim, rn, cex.axis = 0.9, las=3) + axis(2, 1:dim, rn, cex.axis = 0.9, las=1) + text(expand.grid(1:dim, 1:dim), sprintf("%%0.2f", dst), cex=0.6) + dev.off() + """ % ffr(self.pathout['labbe-matrix.png']) + self.add(txt) + self.write() + +class ChronoChi2Script(PrintRScript) : + def make_script(self) : + self.sources([self.analyse.parent.RscriptsPath['Rgraph']]) + print self.parametres + txt = """ + inRData <- "%s" + dendrof <- "%s" + load(inRData) + load(dendrof) + """ % (ffr(self.pathout['RData.RData']), ffr(self.pathout['dendrogramme.RData'])) + txt += """ + svg <- %s + """ % self.parametres['svg'] + txt += """ + tc <- which(grepl("%s",rownames(chistabletot))) + rn <- rownames(chistabletot)[tc] + tc <- tc[order(rn)] + dpt <- chistabletot[tc,] + tot <- afctable[tc,] + tcp <- rowSums(tot) + ptc <- tcp/sum(tcp) + dpt <- t(dpt) + dd <- dpt + """ % self.parametres['var'].replace(u'*', u"\\\\*") + txt += """ + classes <- n1[,ncol(n1)] + tcl <- table(classes) + if ('0' %in% names(tcl)) { + to.vire <- which(names(tcl) == '0') + tcl <- tcl[-to.vire] + } + tclp <- tcl/sum(tcl) + + #chi2 colors + library(ape) + k <- 1e-02 + lcol <- NULL + lk <- k + for (i in 1:5) { + lcol <- c(lcol, qchisq(1-k,1)) + k <- k/10 + lk <- c(lk,k) + } + lcol <- c(3.84, lcol) + lcol <- c(-Inf,lcol) + lcol <- c(lcol, Inf) + lk <- c(0.05,lk) + breaks <- lcol + alphas <- seq(0,1, length.out=length(breaks)) + clod <- rev(as.numeric(tree.cut1$tree.cl$tip.label)) + #end + """ + txt += """ + open_file_graph("%s", w=%i, h=%i, svg=svg) + """ % (ffr(self.parametres['tmpgraph']), self.parametres['width'], self.parametres['height']) + txt += """ + par(mar=c(3,3,3,3)) + mat.graphic <- matrix(c(rep(1,nrow(dd)),c(2:(nrow(dd)+1))), ncol=2) + mat.graphic <- rbind(mat.graphic, c(max(mat.graphic) + 1 , max(mat.graphic) + 2)) + hauteur <- tclp[clod] * 0.9 + heights.graphic <- append(hauteur, 0.1) + layout(mat.graphic, heights=heights.graphic, widths=c(0.15,0.85)) + par(mar=c(0,0,0,0)) + tree.toplot <- tree.cut1$tree.cl + num.label <- as.numeric(tree.cut1$tree.cl$tip.label) + col.tree <- rainbow(length(num.label))[num.label] + tree.toplot$tip.label <- paste('classe ', tree.toplot$tip.label) + plot.phylo(tree.toplot,label.offset=0.1, cex=1.1, no.margin=T, x.lim=20, tip.color = col.tree) + for (i in clod) { + print(i) + par(mar=c(0,0,0,0)) + lcol <- cut(dd[i,], breaks, include.lowest=TRUE) + ulcol <- names(table(lcol)) + lcol <- as.character(lcol) + for (j in 1:length(ulcol)) { + lcol[which(lcol==ulcol[j])] <- j + } + lcol <- as.numeric(lcol) + mcol <- rainbow(nrow(dd))[i] + last.col <- NULL + for (k in alphas) { + last.col <- c(last.col, rgb(r=col2rgb(mcol)[1]/255, g=col2rgb(mcol)[2]/255, b=col2rgb(mcol)[3]/255, a=k)) + } + #print(last.col) + + barplot(rep(1,ncol(dd)), width=ptc, names.arg=FALSE, axes=FALSE, col=last.col[lcol], border=rgb(r=0, g=0, b=0, a=0.3)) + } + plot(0,type='n',axes=FALSE,ann=FALSE) + label.coords <- barplot(rep(1, ncol(dd)), width=ptc, names.arg = F, las=2, axes=F, ylim=c(0,1), plot=T, col='white') + text(x=label.coords, y=0.5, labels=rn[order(rn)], srt=90) + dev.off() + """ + self.add(txt) + self.write() + +class ChronoPropScript(PrintRScript) : + def make_script(self) : + self.sources([self.analyse.parent.RscriptsPath['Rgraph']]) + print self.parametres + txt = """ + inRData <- "%s" + dendrof <- "%s" + load(inRData) + load(dendrof) + """ % (ffr(self.pathout['RData.RData']), ffr(self.pathout['dendrogramme.RData'])) + txt += """ + svg <- %s + """ % self.parametres['svg'] + txt += """ + tc <- which(grepl("%s",rownames(chistabletot))) + rn <- rownames(chistabletot)[tc] + tc <- tc[order(rn)] + dpt <- chistabletot[tc,] + tot <- afctable[tc,] + tcp <- rowSums(tot) + ptc <- tcp/sum(tcp) + dpt <- t(dpt) + dd <- dpt + """ % self.parametres['var'].replace(u'*', u"\\\\*") + txt += """ + classes <- n1[,ncol(n1)] + tcl <- table(classes) + if ('0' %in% names(tcl)) { + to.vire <- which(names(tcl) == '0') + tcl <- tcl[-to.vire] + } + tclp <- tcl/sum(tcl) + """ + txt += """ + open_file_graph("%s", w=%i, h=%i, svg=svg) + """ % (ffr(self.parametres['tmpgraph']), self.parametres['width'], self.parametres['height']) + txt+= """ + ptt <- prop.table(as.matrix(tot), 1) + par(mar=c(10,2,2,2)) + barplot(t(ptt)[as.numeric(tree.cut1$tree.cl$tip.label),], col=rainbow(ncol(ptt))[as.numeric(tree.cut1$tree.cl$tip.label)], width=ptc, las=3, space=0.05, cex.axis=0.7, border=NA) + dev.off() + """ + self.add(txt) + self.write() + +