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])
#
-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")
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")
data2 <- as(data2, "dgCMatrix")
row.names(data2) <- 1:nrow(data2)
""" % DicoPath['TableUc2']
- #log.info('ATTENTION ############# MODEPATATE ####################')
txt += """
- chd1<-CHD(data1, x = nbt, mode.patate = FALSE, 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 += """
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$cuce2
- """ % (DicoPath['uce'], mincl, classif_mode)
+ """ % (classif_mode, mincl, DicoPath['uce'])
txt += """
tree.tot1 <- make_tree_tot(chd1)
fileout.close()
def RPamTxt(corpus, RscriptPath):
- DicoPath = corpus.dictpathout
- param = corpus.parametre
- print param
+ DicoPath = corpus.pathout
+ param = corpus.parametres
txt = """
source("%s")
""" % (RscriptPath['pamtxt'])
save(tree.cut1, file="%s")
open_file_graph("%s", width = 600, height=400)
classes<-n1[,ncol(n1)]
- plot.dendropr(tree_cut1$tree.cl,classes)
+ plot.dendropr(tree_cut1$tree.cl,classes, histo = TRUE)
""" % (DicoPath['Rdendro'],DicoPath['dendro1'])
txt += """
""" % (DictChdTxtOut['afc_facteur'], DictChdTxtOut['afc_col'], DictChdTxtOut['afc_row'])
txt += """
- #xlab <- paste('facteur 1 - ', round(afc$facteur[1,2],2), sep = '')
- #ylab <- paste('facteur 2 - ', round(afc$facteur[2,2],2), sep = '')
- #xlab <- paste(xlab, ' %', sep = '')
- #ylab <- paste(ylab, ' %', sep = '')
- """
-
- txt += """
PARCEX<-%s
- xmin <- min(afc$rowcoord[,1], na.rm = TRUE) + (0.1 * min(afc$rowcoord[,1], na.rm = TRUE))
- xmax <- max(afc$rowcoord[,1], na.rm = TRUE) + (0.1 * max(afc$rowcoord[,1], na.rm = TRUE))
- ymin <- min(afc$rowcoord[,2], na.rm = TRUE) + (0.1 * min(afc$rowcoord[,2], na.rm = TRUE))
- ymax <- max(afc$rowcoord[,2], na.rm = TRUE) + (0.1 * max(afc$rowcoord[,2], na.rm = TRUE))
- print(xmin)
- print(xmax)
- print(ymin)
- print(ymax)
""" % taillecar
txt += """
- PlotAfc2dCoul(afc, as.data.frame(chistabletot), "%s", what='coord', deb=1, fin=(debsup-1), xlab = xlab, ylab = ylab, xmin=xmin, xmax=xmax, ymin = ymin, ymax=ymax)
+ xyminmax <- PlotAfc2dCoul(afc, as.data.frame(chistabletot), "%s", what='coord', deb=1, fin=(debsup-1), xlab = xlab, ylab = ylab)
""" % (DictChdTxtOut['AFC2DL_OUT'])
txt += """
- PlotAfc2dCoul(afc, as.data.frame(chistabletot), "%s", what='coord', deb=debsup, fin=(debet-1), xlab = xlab, ylab = ylab, xmin=xmin, xmax=xmax, ymin = ymin, ymax=ymax)
+ PlotAfc2dCoul(afc, as.data.frame(chistabletot), "%s", what='coord', deb=debsup, fin=(debet-1), xlab = xlab, ylab = ylab, xmin = xyminmax$xminmax[1], xmax = xyminmax$xminmax[2], ymin = xyminmax$yminmax[1], ymax = xyminmax$yminmax[2])
""" % (DictChdTxtOut['AFC2DSL_OUT'])
txt += """
- PlotAfc2dCoul(afc, as.data.frame(chistabletot), "%s", what='coord', deb=debet, fin=fin, xlab = xlab, ylab = ylab, xmin=xmin, xmax=xmax, ymin = ymin, ymax=ymax)
+ if ((fin - debet) > 2) {
+ PlotAfc2dCoul(afc, as.data.frame(chistabletot), "%s", what='coord', deb=debet, fin=fin, xlab = xlab, ylab = ylab, xmin = xyminmax$xminmax[1], xmax = xyminmax$xminmax[2], ymin = xyminmax$yminmax[1], ymax = xyminmax$yminmax[2])
+ }
""" % (DictChdTxtOut['AFC2DEL_OUT'])
txt += """
- PlotAfc2dCoul(afc, as.data.frame(chistabletot), "%s", col=TRUE, what='coord', xlab = xlab, ylab = ylab, xmin=xmin, xmax=xmax, ymin = ymin, ymax=ymax)
+ PlotAfc2dCoul(afc, as.data.frame(chistabletot), "%s", col=TRUE, what='coord', xlab = xlab, ylab = ylab, xmin = xyminmax$xminmax[1], xmax = xyminmax$xminmax[2], ymin = xyminmax$yminmax[1], ymax = xyminmax$yminmax[2])
""" % (DictChdTxtOut['AFC2DCL_OUT'])
# txt += """
# PlotAfc2dCoul(afc, as.data.frame(chistabletot), "%s", what='crl', deb=1, fin=(debsup-1), xlab = xlab, ylab = ylab)
self.packages(['igraph', 'proxy', 'Matrix'])
self.sources([self.analyse.parent.RscriptsPath['simi'], self.analyse.parent.RscriptsPath['Rgraph']])
txt = ''
- if not self.parametres['keep_coord'] :
+ if not self.parametres['keep_coord'] and not self.parametres['type'] == 'simimatrix':
txt += """
dm.path <- "%s"
cn.path <- "%s"
selected.col <- "%s"
""" % (self.pathout['mat01.csv'], self.pathout['actives.csv'], 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=';', quote='"')
+ cn <- read.table(cn.path, sep='\t', quote='"')
colnames(dm) <- cn[,1]
- sel.col <- read.csv2(selected.col)
- dm <- dm[, sel.col[,1] + 1]
+ sel.col <- read.csv2(selected.col, header = FALSE)
+ sel.col <- sel.col[,1] + 1
+ 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' :
+ txt += """
+ dm.path <- "%s"
+ selected.col <- "%s"
+ """ % (self.pathout['mat01.csv'], 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)
+ sel.col <- read.csv2(selected.col, header = FALSE)
+ sel.col <- sel.col[,1] + 1
+ 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 += """
txt += """
mat <- as.matrix(stats::as.dist(mat,diag=TRUE,upper=TRUE))
mat[is.na(mat)] <- 0
- mat[is.infinite(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'
# 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)
+ 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) :
vertex.size <- NULL
"""
else :
- #FIXME
- tmpchi = False
- if tmpchi :
+ if self.parametres['type'] == 'clustersimitxt' :
txt += """
lchi <- read.table("%s")
lchi <- lchi[,1]
- """ % ffr(tmpchi)
- if 'selected_col' in dir(self.tableau) :
- txt += """
- lchi <- lchi[c%s+1]
- """ % datas
- if tmpchi and self.parametres.get('cexfromchi', False) :
+ """ % ffr(self.parametres['tmpchi'])
+ txt += """
+ lchi <- lchi[sel.col]
+ """
+ if self.parametres['type'] == 'clustersimitxt' and self.parametres.get('cexfromchi', False) :
txt += """
label.cex <- norm.vec(lchi, vcexminmax[1], vcexminmax[2])
"""
label.cex <- graph.simi$label.cex
}
"""
- if tmpchi and self.parametres.get('sfromchi', False) :
+ if self.parametres['type'] == 'clustersimitxt' 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 <- NULL
+ vertex.size <- 0
} else {
vertex.size <- graph.simi$eff
}
"""
- txt += """ vertex.size <- NULL """
+ #txt += """ vertex.size <- NULL """
+ if self.parametres['svg'] : svg = 'TRUE'
+ else : svg = 'FALSE'
txt += """
- coords <- plot.simi(graph.simi, p.type='%s',filename="%s", vertex.label = label.v, edge.label = label.e, vertex.col = cols, vertex.label.color = vertex.label.color, vertex.label.cex=label.cex, vertex.size = vertex.size, edge.col = cola, leg=leg, width = width, height = height, alpha = alpha, movie = film)
+ svg <- %s
+ """ % svg
+ txt += """
+ vertex.col <- cols
+ if (!is.null(graph.simi$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, svg = svg)
save.image(file="%s")
""" % (type, self.filename, self.pathout['RData'])
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,])
toprint <- as.matrix(toprint[order(toprint[,1], decreasing=TRUE),])
toprint <- as.matrix(toprint[1:maxword,])
}
- open_file_graph("%s", width = %i, height = %i)
+ 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()