+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()
+ """ % (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()
+
+
+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