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")
row.names(data2) <- 1:nrow(data2)
""" % 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 += """
""" % 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$cuce2
- """ % (DicoPath['uce'], mincl, classif_mode)
+ """ % (classif_mode, mincl, DicoPath['uce'])
txt += """
tree.tot1 <- make_tree_tot(chd1)
save(tree.cut1, file="%s")
classes<-n1[,ncol(n1)]
open_file_graph("%s", width = 600, height=400)
- plot.dendropr(tree.cut1$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()
fileout.close()
def RPamTxt(corpus, RscriptPath):
- DicoPath = corpus.dictpathout
- param = corpus.parametre
+ DicoPath = corpus.pathout
+ param = corpus.parametres
print param
txt = """
source("%s")
chd.result<-Rchdquest("%s","%s","%s", nbt = nbt, mincl = mincl)
n1 <- chd.result$n1
classeuce1 <- chd.result$cuce1
- """ % (DicoPath['Act01'], DicoPath['listeuce1'], DicoPath['uce'])
+ """ % (DicoPath['mat01'], DicoPath['listeuce1'], DicoPath['uce'])
txt += """
tree_tot1 <- make_tree_tot(chd.result$chd)
save(tree.cut1, file="%s")
open_file_graph("%s", width = 600, height=400)
classes<-n1[,ncol(n1)]
- plot.dendropr(tree_cut1$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$facteur[2,2],2), sep = '')
- xlab <- paste(xlab, ' %', sep = '')
- ylab <- paste(ylab, ' %', sep = '')
- """
-
- txt += """
PARCEX<-%s
""" % taillecar
txt += """
- PlotAfc2dCoul(afc, as.data.frame(chistabletot), "%s", what='coord', deb=1, fin=(debsup-1), xlab = xlab, ylab = ylab)
+ xyminmax <- PlotAfc2dCoul(afc, as.data.frame(chistabletot), "%s", what='coord', deb=1, fin=(debsup-1), xlab = xlab, ylab = ylab)
""" % (DictChdTxtOut['AFC2DL_OUT'])
txt += """
- PlotAfc2dCoul(afc, as.data.frame(chistabletot), "%s", what='coord', deb=debsup, fin=(debet-1), xlab = xlab, ylab = ylab)
+ PlotAfc2dCoul(afc, as.data.frame(chistabletot), "%s", what='coord', deb=debsup, fin=(debet-1), xlab = xlab, ylab = ylab, xmin = xyminmax$xminmax[1], xmax = xyminmax$xminmax[2], ymin = xyminmax$yminmax[1], ymax = xyminmax$yminmax[2])
""" % (DictChdTxtOut['AFC2DSL_OUT'])
txt += """
- PlotAfc2dCoul(afc, as.data.frame(chistabletot), "%s", what='coord', deb=debet, fin=fin, xlab = xlab, ylab = ylab)
+ PlotAfc2dCoul(afc, as.data.frame(chistabletot), "%s", what='coord', deb=debet, fin=fin, xlab = xlab, ylab = ylab, xmin = xyminmax$xminmax[1], xmax = xyminmax$xminmax[2], ymin = xyminmax$yminmax[1], ymax = xyminmax$yminmax[2])
""" % (DictChdTxtOut['AFC2DEL_OUT'])
txt += """
- PlotAfc2dCoul(afc, as.data.frame(chistabletot), "%s", col=TRUE, what='coord', xlab = xlab, ylab = ylab)
+ PlotAfc2dCoul(afc, as.data.frame(chistabletot), "%s", col=TRUE, what='coord', xlab = xlab, ylab = ylab, xmin = xyminmax$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)
- 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 += """
+ # 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)
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.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'], \
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):
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")
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)
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))
+ 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))
+ if (length(toinf)) {
+ coordinf <- coord[toinf]
+ valinf <- di[toinf]
+ text(x=coordinf, y=valinf + 0.1, 'i')
+ }
+ 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)]
self.txtgraph = ''
self.packages(['igraph', 'proxy', 'Matrix'])
self.sources([self.analyse.parent.RscriptsPath['simi'], self.analyse.parent.RscriptsPath['Rgraph']])
- txt = """
- dm.path <- "%s"
- cn.path <- "%s"
- selected.col <- "%s"
- """ % (self.pathout['mat01.csv'], self.pathout['actives.csv'], self.pathout['selected.csv'])
- txt += """
- dm <-readMM(dm.path)
- cn <- read.table(cn.path, sep=';', quote='"')
- colnames(dm) <- cn[,1]
- sel.col <- read.csv2(selected.col)
- dm <- dm[, sel.col[,1] + 1]
- """
-
- if self.parametres['coeff'] == 0 :
- method = 'cooc'
+ txt = ''
+ if not self.parametres['keep_coord'] and not self.parametres['type'] == 'simimatrix':
txt += """
- method <- 'cooc'
- mat <- make.a(dm)
+ 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='\t', quote='"')
+ colnames(dm) <- cn[,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)
+ }
"""
- else :
+ 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 += """
+ load("%s")
+ """ % 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 += """
"""
elif self.analyse.indices[self.parametres['coeff']] == 'binomial' :
method = 'binomial'
- txt += """
- method <- 'binomial'
- mat <- binom.sim(dm)
- """
+ 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 += """
- method <-"%s"
- mat <- simil(dm, method = method, diag = TRUE, upper = TRUE, by_rows = FALSE)
- """ % self.analyse.indices[self.parametres['coeff']]
- txt += """
- mat <- as.matrix(stats::as.dist(mat,diag=TRUE,upper=TRUE))
- mat[is.na(mat)] <- 0
- mat[is.infinite(mat)] <- 0
- """
+ mat <- as.matrix(stats::as.dist(mat,diag=TRUE,upper=TRUE))
+ mat[is.na(mat)] <- 0
+ mat[is.infinite(mat)] <- 0
+ """
+ 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 :
txt += """
et <- list()
"""
- for i,et in enumerate(self.tableau.etline) :
+ for i, line in enumerate(self.parametres['listet']) :
txt+= """
et[[%i]] <- c(%s)
- """ % (i+1, ','.join(et[1:]))
+ """ % (i+1, ','.join([`val + 1` for val in line]))
txt+= """
unetoile <- c('%s')
- """ % ("','".join([val[0] for val in self.tableau.etline]))
+ """ % ("','".join([val for val in self.parametres['selectedstars']]))
txt += """
fsum <- NULL
rs <- rowSums(dm)
cols <- vertex.label.color
chivertex.size <- norm.vec(toblack, vcexminmax[1], vcexminmax[2])
- """ % (self.parent.RscriptsPath['chdfunct'])
+ """ % (self.analyse.parent.RscriptsPath['chdfunct'])
else :
txt += """
vertex.label.color <- 'black'
# 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])
"""
}
"""
txt += """ vertex.size <- NULL """
+ if self.parametres['svg'] : svg = 'TRUE'
+ else : svg = 'FALSE'
+ txt += """
+ svg <- %s
+ """ % svg
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)
+ 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 = svg)
save.image(file="%s")
""" % (type, self.filename, self.pathout['RData'])
projects / iramuteq / blobdiff