Skip to content
Extraits de code Groupes Projets
Valider beec5db3 rédigé par Alain Guillet's avatar Alain Guillet
Parcourir les fichiers

Fichier server.R

parent 0bf75fa7
Aucune branche associée trouvée
Aucune étiquette associée trouvée
Aucune requête de fusion associée trouvée
Masquer les modifications d'espaces
En ligne Côte à côte
Affichage de
avec 638 ajouts et 0 suppression
server.R 0 → 100644
+ 638
− 0
Voir le fichier @ beec5db3
## simpleIC Shiny/R app server.R
##
## Author(s) :
## -----------
## J.J.
## Orginal version by Grégoire Vincke http://www.uclouvain.be/gregoire.vincke
## For Statistical eLearning Tools http://sites.uclouvain.be/selt/
##
## Licences :
## ---------
## CC-BY for the web page http://sites.uclouvain.be/selt/shiny/testhypic
## see http://creativecommons.org/licenses/by/2.0/be/ for more informations
##
## GPLv2 for source code on https://github.com/uclouvain-selt/shiny
## See LICENCE.tx or http://www.gnu.org/licenses/old-licenses/gpl-2.0.html for more informations
Sys.setlocale("LC_ALL", "fr_FR.UTF-8")#to be sure that accents in text will be allowed in plots
library(shiny)
# library(plotrix)
# library(xtable)
# library(ggplot2)
debug<-0
color.blue<-rgb(0,0,0.9)
color.true<-rgb(0,0.7,0)
color.false<-rgb(1,0,0,0.9)
oui.color.true<-rgb(0.3,0.3,0.3)
oui.color.false<-rgb(.6,.6,.6)
text.color.true<-rgb(0.2,0.2,0.2)
text.color.false<-rgb(.5,.5,.5)
non.color.true<-rgb(0.8,0.8,0.8)
non.color.false<-rgb(0.93,0.93,0.93)
density.true<-10
density.false<-25
y.delta<-0.1 #factor to set delta between rows of datas in plots
hypoth.text.levels<-c(1,0.7,0.4,0.1)
x.lim.min<-0
x.lim.max<-60
x.amp<-x.lim.max-x.lim.min
# possible values for the mean, cf ui.R mx1, mx0, mx
mu.vec<-c(x.lim.min:x.lim.max)#c(25:75)
shinyServer(function(input, output,session){
rv <- reactiveValues()# Create a reactiveValues object, to let us use settable reactive values
rv$last.takesample.value<-0
#rv$samples.z<-list() # all observations
rv$samples.mat<-c() # matrix of all observations, each line one sample
rv$new.sample<-c() # new matrix of observations, each line one sample
rv$cv.ls<-list() # calculated values
rv$lastAction <- 'none' # To start out, lastAction == NULL, meaning nothing clicked yet
# An observe block for each button, to record that the action happened
# Calculations only needed if one of these values are changed, so observe them
rv$mx.c<-0 # population mean
rv$sx.c<-0 # population deviation
rv$IC.k.c<-0 # IC length
rv$typIC.c<-'' # Type 'eCVk' empiric, 'vCVk' variance connue, 'sCVk' variance inconnue
# Create all samples new if a change is made in sample size
rv$n.c<-0 # sample size as in ui
rv$tn.c<-0 # total numer of samples
observe({
if(input$visM){
js_string <- '$(".span8").width(500);'
session$sendCustomMessage(type='jsCode', list(value = js_string))
}
})
# if take sample
observe({
if (input$takesample != 0) {
rv$lastAction <- 'takesample'
}
})
# if reset all new
observe({
if(input$reset !=0){
rv$lastAction <- 'reset'
rv$last.takesample.value<-0
rv$samples.mat<-c()
rv$cv.ls<-list()
rv$mx.c<-0 # population mean
rv$sx.c<-0 # population deviation
rv$IC.k.c<-0 # IC length
rv$typIC.c<-'' # Type 'eCVk' empiric, 'vCVk' variance connue, 'sCVk' variance inconnue
rv$n.c<-0 # sample size as in ui
rv$tn.c<-0 # total numer of samples
updateSliderInput(session, "mx1",value = sample(c(31:35),1))
updateSliderInput(session, "sx",value = sample(seq(from = 2, to = 3.5, by = 0.5),1))
updateCheckboxInput(session, "muKn",value=FALSE)
updateCheckboxInput(session, "sigKn",value=FALSE)
}
})
## observe({
## if (input$n != rv$n.c) {
## isolate({#Now do the expensive stuff
## rv$n.c<-input$n
## rv$samples.mat<-matrix(rnorm((rv$tn.c+input$ns)*rv$n.c),ncol=rv$n.c)
## rv$tn.c<-length(rv$samples.mat[,1]) # new total number of samples
## })
## }
## })
getSamples<-reactive({#créee n valeurs aléatoires N(0;1) quand input$takesample est implémenté (quand le bouton takesample est pressé)
if(input$takesample > rv$last.takesample.value && rv$lastAction == "takesample"){
return(isolate({#Now do the expensive stuff
rv$new.sample<-matrix(rnorm(input$ns*input$n),ncol=input$n)
return(TRUE)
}))
} else {
return(FALSE)
}
})
getPlotHeight <- function() {
if(input$display=="default") {
unit.height<-250 #cannot be auto because height is already "auto" in ui and double auto = conflict
}
if(input$display=="1024") {
unit.height<-180
}
if(input$display=="800") {
unit.height<-140 #160 for real full page
}
return(3*unit.height)
}
getPlotWidth <- function() {
if(input$display=="default") {
full.plot.width<-1310-310#"auto"
}
if(input$display=="1024") {
full.plot.width<-900-310
}
if(input$display=="800") {
full.plot.width<-700-310
}
if(input$visM && input$display!="default"){
full.plot.width<-full.plot.width+310
}
return(full.plot.width)
}
getInputValues<-reactive({
return(input)#collect all inputs
})
getComputedValues<-reactive({
# returns TRUE if new calculations other wise FALSE
# results hold in rv$cv.ls
# gives TRUE back if new values calculated otherwise FALSE
calc.new<-FALSE # we do not want to calculate all values again
# did we create new samples?
sample.new<-getSamples() # if TRUE then append new observations from rv$new.sample
v<-getInputValues() # get all values of input list
# check if sample size was changed
if (v$n != rv$n.c){# if changed create a new observation matrix of correct size
rv$n.c<-v$n
if(rv$tn.c>0){
rv$samples.mat<-matrix(rnorm(rv$tn.c*rv$n.c),ncol=rv$n.c)
}
calc.new<-TRUE # we have to calculate all values again
}
if(sample.new){#if new observations created, append them
sample.mat<-mat.or.vec(rv$tn.c + v$ns,v$n)
if(rv$tn.c>0){
sample.mat[1:rv$tn.c,]<-rv$samples.mat
}
sample.mat[(rv$tn.c+1):(rv$tn.c+v$ns),]<-rv$new.sample
rv$samples.mat<-sample.mat # new observations
rv$tn.c<-length(rv$samples.mat[,1]) # new total number of samples
calc.new<-TRUE # we have to calculate all values again
}
# check if caluations are needed due to parameter changes
# could be still optimized since a change in rv$IC.k.c does not needs a whole new calcul
if (v$mx1 != rv$mx.c){# population mean changed
rv$mx.c<-v$mx1 # update
calc.new<-TRUE
}
if (v$sx != rv$sx.c){# population sd changed
rv$sx.c<-v$sx # update
calc.new<-TRUE
}
if (v$k != rv$IC.k.c){# IC length changed
rv$IC.k.c<-v$k # update
calc.new<-TRUE
}
if (v$CVk != rv$typIC.c){# type IC changed
rv$typIC.c<-v$CVk # update
calc.new<-TRUE
}
# new calulations if new obsarvations or
if(calc.new){
cv<-list()#created empty computed values list
## Define reality parameters
cv$vx<-v$sx^2#compute variance of Reality distribution
## Computation of x y coordinates for Normal curve of Reality
z<-seq(-5,5,length=100)
cv$xr<-(z*v$sx)+v$mx1 #x for Reality
cv$yr<-dnorm(cv$xr,mean=v$mx1,sd=v$sx)#y for Reality
## Computation of sample related values ##
cv$samples.x.mat<-c() # matrix of observations, each line a sample
cv$samples.x.m.vec<-c() # vector of mean values, each line a sample
cv$samples.x.sd.vec<-c() # vector of sd values, each line a sample
cv$ic.k.limit.mat<-c() # matrix of limits, columns lower and upper bound , lines by sample
cv$ic.k.inc.allmu.mat<-c() # matrix of TRUE/FALSE if mu in IC columns all mu.vec=c(20:60) and lines by sample
cv$pc.ic.k.inc.allmu.vec<-c() # for all mu increment percentage covered by IC
cv$n.ic.k.inc.allmu.vec<-c() # for all mu increment number covered by IC
cv$n.samples<-rv$tn.c # number of samples
cv$samples.x.n.toshow<-0
if(cv$n.samples>0){
cv$samples.x.mat<-mat.or.vec(cv$n.samples,v$n)
cv$ic.k.limit.mat<-mat.or.vec(cv$n.samples,2)
cv$vect.n.samples<-c(1:cv$n.samples)
cv$samples.x.mat<-round((rv$samples.mat*v$sx)+v$mx1,2)#Then sample values are compute with mx1 mean and standard deviation
## Computation of descriptives
cv$samples.x.m.vec<-round(apply(cv$samples.x.mat,1,mean),2)#means of samples
cv$samples.x.sd.vec<-round(apply(cv$samples.x.mat,1,sd),2)#sds of samples
## Computation of confidence intervals for the mean µ ##
if(v$CVk == 'vCVk'){#compute the CI limits with k value and known variance
cv$ic.k.limit.mat[,1]<-round(cv$samples.x.m.vec-v$k*v$sx*(v$n)^(-.5),2)
cv$ic.k.limit.mat[,2]<-round(cv$samples.x.m.vec+v$k*v$sx*(v$n)^(-.5),2)
}
if(v$CVk == 'sCVk'){#compute the CI limits with k value and unknown variance
cv$ic.k.limit.mat[,1]<-round(cv$samples.x.m.vec-v$k*cv$samples.x.sd.vec*(v$n)^(-.5),2)
cv$ic.k.limit.mat[,2]<-round(cv$samples.x.m.vec+v$k*cv$samples.x.sd.vec*(v$n)^(-.5),2)
}
if(v$CVk == 'eCVk'){#compute the CI limits with empiric k value
cv$ic.k.limit.mat[,1]<-round(cv$samples.x.m.vec-v$k,2)
cv$ic.k.limit.mat[,2]<-round(cv$samples.x.m.vec+v$k,2)
}
## Check for all values in mu.vec if in IC
cv$ic.k.inc.allmu.mat<-sapply(mu.vec,function(x){return (cv$ic.k.limit.mat[,1] <=x & x<=cv$ic.k.limit.mat[,2])})
## Calculate for all values in mu.vec frequencies absolute and relative
cv$n.ic.k.inc.allmu.vec<-apply(matrix(cv$ic.k.inc.allmu.mat,ncol=length(mu.vec)),2,sum)
cv$pc.ic.k.inc.allmu.vec<-round(cv$n.ic.k.inc.allmu.vec/cv$n.samples,3)*100
## Define colors
cv$ic.k.inc.mu.color.mat<-matrix(ifelse(cv$ic.k.inc.allmu.mat,oui.color.true,oui.color.false),ncol=length(mu.vec))
cv$ic.k.inc.mu0.color.mat<-matrix(ifelse(cv$ic.k.inc.allmu.mat,color.false,color.true),ncol=length(mu.vec))
cv$ic.k.inc.mu1.color.mat<-matrix(ifelse(cv$ic.k.inc.allmu.mat,color.true,color.false),ncol=length(mu.vec))
## Define subset to plot
cv$samples.x.from<-1
if(cv$n.samples>v$nss){
cv$samples.x.from<-cv$n.samples-v$nss+1
}
cv$samples.x.to<-cv$n.samples
cv$samples.x.mat.toshow<-matrix(cv$samples.x.mat[cv$samples.x.from:cv$samples.x.to,],ncol=v$n)
cv$samples.x.m.vec.toshow<-cv$samples.x.m.vec[cv$samples.x.from:cv$samples.x.to]
cv$samples.x.sd.vec.toshow<-cv$samples.x.sd.vec[cv$samples.x.from:cv$samples.x.to]
cv$samples.x.i.vec.toshow<-c(cv$samples.x.from:cv$samples.x.to)
cv$ic.k.limit.mat.toshow<-matrix(cv$ic.k.limit.mat[cv$samples.x.from:cv$samples.x.to,],ncol=2)
cv$samples.y.mat.toshow<-c() # plot line by line the values, here corresponding y-values
cv$samples.x.n.toshow<-length(cv$samples.x.mat.toshow[,1])
cv$ic.k.inc.mu.color.vec.toshow<-c() # color the IC for mu
cv$ic.k.inc.mu0.color.vec.toshow<-c() # color the IC for mu0
cv$ic.k.inc.mu1.color.vec.toshow<-c() # color the IC for mu1
if(cv$samples.x.n.toshow>0){
cv$samples.y.mat.toshow<-matrix(rep(y.delta/(v$nss+1)*c(1:cv$samples.x.n.toshow),length(cv$samples.x.mat.toshow[1,])),nrow=length(cv$samples.x.mat.toshow[,1]))
## ## Define colors if IC covers µ or µ0 or µ1
cv$ic.k.inc.mu.color.vec.toshow<-cv$ic.k.inc.mu.color.mat[cv$samples.x.from:cv$samples.x.to,v$mx-mu.vec[1]+1]
cv$ic.k.inc.mu0.color.vec.toshow<-cv$ic.k.inc.mu0.color.mat[cv$samples.x.from:cv$samples.x.to,v$mx0-mu.vec[1]+1]
cv$ic.k.inc.mu1.color.vec.toshow<-cv$ic.k.inc.mu1.color.mat[cv$samples.x.from:cv$samples.x.to,v$mx1-mu.vec[1]+1]
}
}
rv$cv.ls<-cv # set new values
}
## Last takesample value
rv$last.takesample.value<-v$takesample
return(calc.new)
})
output$plotEmp <- renderPlot({
v<-getInputValues()
calc.new<-getComputedValues() # TRUE if new values have been calculated
cv<-rv$cv.ls # holds calculated values
# if graphic values of parameter are changed without new calculus, so here calculate new values
if(v$display=="default") {
cex.samples<-2.2 #size of text describing samples (2.2)
cex.param<-3.5 #size of text of parameter µ µ'', etc (3.5)
cex.title<-2.2
y.delta<-0.1 #factor to set delta between rows of datas in plots
ic.bar.half.height<-0.004
}
if(v$display=="1024") {
cex.samples<-1.7 #size of text describing samples (2.2)
cex.param<-2.5 #size of text of parameter µ µ'', etc (3.5)
cex.title<-1.7
y.delta<-0.1 #factor to set delta between rows of datas in plots
ic.bar.half.height<-0.004
}
if(v$display=="800") {
cex.samples<-1.5 #size of text describing samples (2.2)
cex.param<-2 #size of text of parameter µ µ'', etc (3.5)
cex.title<-1.7
y.delta<-0.1 #factor to set delta between rows of datas in plots
ic.bar.half.height<-0.004
}
## Define subset to plot
if(cv$n.samples>0){
cv$samples.x.from<-1
if(cv$n.samples>v$nss){
cv$samples.x.from<-cv$n.samples-v$nss+1
}
cv$samples.x.to<-cv$n.samples
cv$samples.x.mat.toshow<-matrix(cv$samples.x.mat[cv$samples.x.from:cv$samples.x.to,],ncol=v$n)
cv$samples.x.m.vec.toshow<-cv$samples.x.m.vec[cv$samples.x.from:cv$samples.x.to]
cv$samples.x.sd.vec.toshow<-cv$samples.x.sd.vec[cv$samples.x.from:cv$samples.x.to]
cv$samples.x.i.vec.toshow<-c(cv$samples.x.from:cv$samples.x.to)
cv$ic.k.limit.mat.toshow<-matrix(cv$ic.k.limit.mat[cv$samples.x.from:cv$samples.x.to,],ncol=2)
cv$samples.y.mat.toshow<-c() # plot line by line the values, here corresponding y-values
cv$samples.x.n.toshow<-length(cv$samples.x.mat.toshow[,1])
cv$ic.k.inc.mu.color.vec.toshow<-c() # color the IC for mu
cv$ic.k.inc.mu0.color.vec.toshow<-c() # color the IC for mu0
cv$ic.k.inc.mu1.color.vec.toshow<-c() # color the IC for mu1
if(cv$samples.x.n.toshow>0){
cv$samples.y.mat.toshow<-matrix(rep(y.delta/(v$nss+1)*c(1:cv$samples.x.n.toshow),length(cv$samples.x.mat.toshow[1,])),nrow=length(cv$samples.x.mat.toshow[,1]))
## ## Define colors if IC covers µ or µ0 or µ1
cv$ic.k.inc.mu.color.vec.toshow<-cv$ic.k.inc.mu.color.mat[cv$samples.x.from:cv$samples.x.to,v$mx-mu.vec[1]+1]
cv$ic.k.inc.mu0.color.vec.toshow<-cv$ic.k.inc.mu0.color.mat[cv$samples.x.from:cv$samples.x.to,v$mx0-mu.vec[1]+1]
cv$ic.k.inc.mu1.color.vec.toshow<-cv$ic.k.inc.mu1.color.mat[cv$samples.x.from:cv$samples.x.to,v$mx1-mu.vec[1]+1]
}
}
m<-matrix(c(1,2,3,4,5,6),3,2,byrow=TRUE)
layout(m,width=c(3,1))
##-------------------------------------------
## Plot always Reality ##
##-------------------------------------------
cv$maxdmx=0.05
par(mai=c(0.3,0.6,0.5,0))
label<-""
if(v$showreality){
label<-"Density"
}
plot(c(0),c(-5),lty=1,lwd=1,col="black",yaxt="n",bty="n",las=1,xaxs="i",yaxs="i",cex.lab=1,cex.axis=1.5,xlim=c(x.lim.min,x.lim.max),ylim=c(0,cv$maxdmx*2.1),xlab="",ylab=label,xaxp=c(x.lim.min,x.lim.max,20),main=bquote(paste("Echantillons prélevés :")),cex.main=cex.title)
if(debug){
box(which="outer",lty = 'dotted', col = 'red')
box(which="figure",lty = 'dotted', col = 'blue')
box(which="plot",lty = 'dotted', col = 'blue')
}
if(cv$samples.x.n.toshow>0){
for(i in 1:cv$samples.x.n.toshow){
points(cv$samples.x.mat.toshow[i,],cv$samples.y.mat.toshow[i,],cex=cex.samples*0.8)
text(cv$samples.x.m.vec.toshow[i],cv$samples.y.mat.toshow[i,1],labels=bquote(bar(x)[.(cv$samples.x.i.vec.toshow[i])]),cex=cex.samples*1.2,col=color.blue)
}
}
if(v$showreality){
axis(2,las=2,yaxp=c(0,signif(cv$maxdmx,1),5),cex.axis=1.2)
points(cv$xr,cv$yr,type="l")
text(1,signif(cv$maxdmx,1)*0.75,labels=bquote(paste(N *"~"* ( mu *","* sigma^2 ) ," ", N *"~"* (.(v$mx1)*","*.(cv$vx)) ,sep='')),cex=1.4, pos=4)
}
if(v$muKn){
## Plot true mean only if known
lines(x<-c(v$mx1,v$mx1),y <- c(0,cv$maxdmx*1.8),lty=1,lwd=2)
text(v$mx1,cv$maxdmx*1.95,labels=bquote(mu),cex=cex.param,col=color.blue)
}
## empty plot for layout
par(mai=c(0.3,0,0.5,0))
plot(c(0,1),c(0,0),col="white",xaxt="n",yaxt="n",xlab="",ylab="",ylim=c(0,cv$maxdmx*2.1),bty="n",las=1)
if(debug){
box(which="figure",lty = 'dotted', col = 'blue')
box(which="plot",lty = 'dotted', col = 'blue')
}
if(v$empPl){
mtext(bquote(paste("Descriptives : ", N == .(cv$n.samples), sep="")),side=1,line=1,at=0.05,adj=0)
if(cv$samples.x.n.toshow>0){
for(i in 1:cv$samples.x.n.toshow){
text(0,cv$samples.y.mat.toshow[i,1],labels=bquote(paste(bar(x)[.(cv$samples.x.i.vec.toshow[i])] == .(sprintf("%.2f",cv$samples.x.m.vec.toshow[i])),sep="")),cex=cex.samples,col=color.blue,pos=4)
text(0.5,cv$samples.y.mat.toshow[i,1],labels=bquote(paste(s[.(cv$samples.x.i.vec.toshow[i])] == .(sprintf("%.2f",cv$samples.x.sd.vec.toshow[i])),sep="")),cex=cex.samples,pos=4)
}
}
}
if(v$icPl){
##-------------------------------------------
## Plot IC ##
##-------------------------------------------
cv$maxdmx=0.05
par(mai=c(0.3,0.6,0.5,0))
plot(c(0),c(-5),lty=1,lwd=1,col="black",yaxt="n",bty="n",las=1,xaxs="i",yaxs="i",cex.lab=1,cex.axis=1.5,xlim=c(x.lim.min,x.lim.max),ylim=c(0,cv$maxdmx*2.1),ylab="",xlab="",xaxp=c(x.lim.min,x.lim.max,20),main=bquote(paste("Intervalles de confiance:")),cex.main=cex.title)
if(debug){
box(which="figure",lty = 'dotted', col = 'blue')
box(which="plot",lty = 'dotted', col = 'blue')
}
if(v$cvPl != "non"){
if(v$cvPl == "oui"){
## Plot mean mx
lines(x<-c(v$mx,v$mx),y <- c(0,cv$maxdmx*1.8),lty=1,lwd=3,col=oui.color.true)
text(v$mx,cv$maxdmx*1.95,labels=v$mx,cex=cex.param*0.75,col=oui.color.true)
help.color.vec<-cv$ic.k.inc.mu.color.vec.toshow
}
if(v$cvPl == "parOri" || (v$cvPl == "parAlt" && v$mx1 == v$mx0)){
## Plot mean mx1
lines(x<-c(v$mx1,v$mx1),y <- c(0,cv$maxdmx*1.8),lty=1,lwd=3,col=color.true)
text(v$mx1,cv$maxdmx*1.95,labels=bquote(mu),cex=cex.param,col=color.true)
help.color.vec<-cv$ic.k.inc.mu1.color.vec.toshow
}
if(v$cvPl == "parAlt" && v$mx1 != v$mx0){
lines(x<-c(v$mx0,v$mx0),y <- c(0,cv$maxdmx*1.8),lty=1,lwd=3,col=color.false)
text(v$mx0,cv$maxdmx*1.95,labels=bquote(paste(mu,"''",sep="")),cex=cex.param,col=color.false)
help.color.vec<-cv$ic.k.inc.mu0.color.vec.toshow
}
} else {
help.color.vec<-ifelse(cv$ic.k.inc.mu0.color.vec.toshow,"black","black")
}
if(cv$samples.x.n.toshow>0){
for(i in 1:cv$samples.x.n.toshow){
polygon(c(cv$ic.k.limit.mat.toshow[i,1],cv$ic.k.limit.mat.toshow[i,1],cv$ic.k.limit.mat.toshow[i,2],cv$ic.k.limit.mat.toshow[i,2]),c(cv$samples.y.mat.toshow[i,1]-ic.bar.half.height,cv$samples.y.mat.toshow[i,1]+ic.bar.half.height,cv$samples.y.mat.toshow[i,1]+ic.bar.half.height,cv$samples.y.mat.toshow[i,1]-ic.bar.half.height),col=help.color.vec[i])
text(cv$samples.x.m.vec.toshow[i],cv$samples.y.mat.toshow[i,1],labels=bquote(bar(x)[.(cv$samples.x.i.vec.toshow[i])]),cex=cex.samples*1.2,col="blue")
lines(x<-c(cv$ic.k.limit.mat.toshow[i,1],cv$samples.x.m.vec.toshow[i]-1),y <- c(cv$samples.y.mat.toshow[i,1],cv$samples.y.mat.toshow[i,1]),lwd=1,lty=2,col="black")
lines(x<-c(cv$samples.x.m.vec.toshow[i]+1,cv$ic.k.limit.mat.toshow[i,2]),y <- c(cv$samples.y.mat.toshow[i,1],cv$samples.y.mat.toshow[i,1]),lwd=1,lty=2,col="black")
}
}
## empty plot for layout
par(mai=c(0.3,0,0.5,0))
plot(c(0,1),c(0,0),col="white",xaxt="n",yaxt="n",xlab="",ylab="",ylim=c(0,cv$maxdmx*2.1),bty="n",las=1)
if(debug){
box(which="figure",lty = 'dotted', col = 'blue')
box(which="plot",lty = 'dotted', col = 'blue')
}
if(v$empPl){
mtext(bquote(paste("Intervalles : ", N == .(cv$n.samples), sep="")),side=1,line=1,at=0.05,adj=0)
if(cv$samples.x.n.toshow>0){
for(i in 1:cv$samples.x.n.toshow){
if(v$cvPl != "non"){
if(v$cvPl == "oui"){
help.color.vec<-cv$ic.k.inc.mu.color.vec.toshow
}
if(v$cvPl == "parOri" || (v$cvPl == "parAlt" && v$mx1 == v$mx0)){
## Plot mean mx1
help.color.vec<-cv$ic.k.inc.mu1.color.vec.toshow
}
if(v$cvPl == "parAlt" && v$mx1 != v$mx0){
help.color.vec<-cv$ic.k.inc.mu0.color.vec.toshow
}
} else {
help.color.vec<-ifelse(cv$ic.k.inc.mu0.color.vec.toshow,"black","black")
}
if(v$thresholds == "formula"){
if(v$CVk == 'eCVk'){
text(0,cv$samples.y.mat.toshow[i,1],bquote(paste(group("[",list(bar(x)[.(cv$samples.x.i.vec.toshow[i])]-c,bar(x)[.(cv$samples.x.i.vec.toshow[i])]+c),"]"),sep="")),las=2,col=help.color.vec[i],cex=cex.samples,pos=4)
}
if(v$CVk == 'vCVk'){
text(0,cv$samples.y.mat.toshow[i,1],bquote(paste(group("[",list(bar(x)[.(cv$samples.x.i.vec.toshow[i])]-c*'*'*sigma/sqrt(n),bar(x)[.(cv$samples.x.i.vec.toshow[i])]+c*'*'*sigma/sqrt(n)),"]"),sep="")),las=2,col=help.color.vec[i],cex=cex.samples,pos=4)
}
if(v$CVk == 'sCVk'){
text(0,cv$samples.y.mat.toshow[i,1],bquote(paste(group("[",list(bar(x)[.(cv$samples.x.i.vec.toshow[i])]-c*'*'*s[.(cv$samples.x.i.vec.toshow[i])]/sqrt(n),bar(x)[.(cv$samples.x.i.vec.toshow[i])]+c*'*'*s[.(cv$samples.x.i.vec.toshow[i])]/sqrt(n)),"]"),sep="")),las=2,col=help.color.vec[i],cex=cex.samples,pos=4)
}
}
if(v$thresholds == "calcul"){
if(v$CVk == 'eCVk'){
text(0,cv$samples.y.mat.toshow[i,1],bquote(paste(group("[",list(.(sprintf("%.2f",cv$samples.x.m.vec.toshow[i]))-.(v$k),.(sprintf("%.2f",cv$samples.x.m.vec.toshow[i]))+.(v$k)),"]"),sep="")),las=2,col=help.color.vec[i],cex=cex.samples,pos=4)
}
if(v$CVk == 'vCVk'){
text(0,cv$samples.y.mat.toshow[i,1],bquote(paste(group("[",list(.(sprintf("%.1f",cv$samples.x.m.vec.toshow[i]))-.(v$k)*'*'*.(v$sx)/.(sprintf("%.1f",v$n^.5)),.(sprintf("%.1f",cv$samples.x.m.vec.toshow[i]))+.(v$k)*'*'*.(v$sx)/.(sprintf("%.1f",v$n^.5))),"]"),sep="")),las=2,col=help.color.vec[i],cex=cex.samples,pos=4)
}
if(v$CVk == 'sCVk'){
text(0,cv$samples.y.mat.toshow[i,1],bquote(paste(group("[",list(.(sprintf("%.1f",cv$samples.x.m.vec.toshow[i]))-.(v$k)*'*'*.(sprintf("%.1f",cv$samples.x.sd.vec.toshow[i]))/.(sprintf("%.1f",v$n^.5)),.(sprintf("%.1f",cv$samples.x.m.vec.toshow[i]))+.(v$k)*'*'*.(sprintf("%.1f",cv$samples.x.sd.vec.toshow[i]))/.(sprintf("%.1f",v$n^.5))),"]"),sep="")),las=2,col=help.color.vec[i],cex=cex.samples,pos=4)
}
}
if(v$thresholds == "result"){
text(0,cv$samples.y.mat.toshow[i,1],bquote(paste(group("[",list(.(sprintf("%.2f",cv$ic.k.limit.mat.toshow[i,1])),.(sprintf("%.2f",cv$ic.k.limit.mat.toshow[i,2]))),"]"),sep="")),las=2,col=help.color.vec[i],cex=cex.samples,pos=4)
}
}
}
}
if(v$cvPl != "non" && v$freqPl == "freqPloui"){
par(mai=c(0.3,0.6,0.5,0))
## Plot bar plot of includes 2 class %
if(cv$n.samples>0){
includes<-t(matrix(c(100-cv$pc.ic.k.inc.allmu.vec,cv$pc.ic.k.inc.allmu.vec),ncol=2))
} else {
includes<-t(matrix(c(rep(0,length(mu.vec)),100-rep(0,length(mu.vec))),ncol=2))
}
barplot.kH1<-barplot(includes,names.arg=mu.vec,ylim=c(0,100),col=c(non.color.false,non.color.true),cex.names=1.25,cex.axis=1.5,beside=FALSE,xaxs="i",space=0,yaxt="n",las=2)
axis(2,las=2,yaxp=c(0,100,2),cex.axis=1.5)#to have las=2 for horizontal labels on y axis
mtext("%",side=2,line=3,at=50)
title(main=bquote(paste("% de couverture par les intervalles de confiance pour ",sep=" ")),adj=1,cex.main=cex.title)
if(debug){
box(which="figure",lty = 'dotted', col = 'blue')
box(which="plot",lty = 'dotted', col = 'blue')
}
# barplot.kH1 is the vector of positions of th bars which we use next
if(v$cvPl == "oui" && cv$n.samples>0){
barplot.spp<-barplot(matrix(c(100-cv$pc.ic.k.inc.allmu.vec[(v$mx-mu.vec[1]+1)],cv$pc.ic.k.inc.allmu.vec[(v$mx-mu.vec[1]+1)]),ncol=1),col=c(oui.color.false,oui.color.true), add=TRUE,beside=FALSE,space=(barplot.kH1[(v$mx-mu.vec[1]+1)]-0.5),axes=FALSE)
}
if(v$cvPl == "parOri" && cv$n.samples>0 || (v$cvPl == "parAlt" && v$mx1 == v$mx0)){
barplot.spp<-barplot(matrix(c(100-cv$pc.ic.k.inc.allmu.vec[(v$mx1-mu.vec[1]+1)],cv$pc.ic.k.inc.allmu.vec[(v$mx1-mu.vec[1]+1)]),ncol=1),col=c(color.false,color.true), add=TRUE,beside=FALSE,space=(barplot.kH1[(v$mx1-mu.vec[1]+1)]-0.5),axes=FALSE)
}
if(v$cvPl == "parAlt" && cv$n.samples>0 && v$mx1 != v$mx0){
barplot.spp<-barplot(matrix(c(100-cv$pc.ic.k.inc.allmu.vec[(v$mx0-mu.vec[1]+1)],cv$pc.ic.k.inc.allmu.vec[(v$mx0-mu.vec[1]+1)]),ncol=1),col=c(color.true,color.false), add=TRUE,beside=FALSE,space=(barplot.kH1[(v$mx0-mu.vec[1]+1)]-0.5),axes=FALSE)
}
## empty plot for layout
par(mai=c(0.3,0,0.5,0))
plot(c(0,1),c(0,0),col="white",xaxt="n",yaxt="n",xlab="",ylab="",ylim=c(0,cv$maxdmx*2.1),bty="n",las=1)
if(debug){
box(which="figure",lty = 'dotted', col = 'blue')
box(which="plot",lty = 'dotted', col = 'blue')
}
if(v$cvPl == "oui"){
title(main=bquote(paste("la valeur ",.(v$mx)," :",sep=" ")),adj=0,cex.main=cex.title)
y.max<-cv$maxdmx*2.1
text(0.425,y.max*0.8,bquote(paste("IC couvrent ",.(v$mx)," ? ",sep=" ")),cex=cex.samples)
text(0.425,y.max*0.6,bquote(paste("n",sep=" ")),cex=cex.samples)
text(0.675,y.max*0.6,bquote(paste("%",sep=" ")),cex=cex.samples)
text(0.175,y.max*0.4,"Oui",col=text.color.true,cex=cex.samples*0.85)
text(0.175,y.max*0.2,"Non",col=text.color.false,cex=cex.samples*0.85)
if(cv$n.samples>0){
# barplot.spp<-barplot(matrix(c(100-cv$pc.ic.k.inc.allmu.vec[(v$mx-mu.vec[1]+1)],cv$pc.ic.k.inc.allmu.vec[(v$mx-mu.vec[1]+1)]),ncol=1),col=c(oui.color.false,oui.color.true), add=TRUE,beside=FALSE,space=(barplot.kH1[(v$mx-mu.vec[1]+1)]-0.5),axes=FALSE)
ICvsmu0.mat<-matrix(c(cv$n.ic.k.inc.allmu.vec[(v$mx-mu.vec[1]+1)],cv$n.samples-cv$n.ic.k.inc.allmu.vec[(v$mx-mu.vec[1]+1)],cv$pc.ic.k.inc.allmu.vec[(v$mx-mu.vec[1]+1)],100-cv$pc.ic.k.inc.allmu.vec[(v$mx-mu.vec[1]+1)]),ncol=2)
ICvsmu0.mat<-round(ICvsmu0.mat,0)
text(0.425,y.max*0.4,bquote(paste(.(ICvsmu0.mat[1,1]),sep=" ")),col=text.color.true,cex=cex.samples)
text(0.675,y.max*0.4,bquote(paste(.(ICvsmu0.mat[1,2]),sep=" ")),col=text.color.true,cex=cex.samples)
text(0.425,y.max*0.2,bquote(paste(.(ICvsmu0.mat[2,1]),sep=" ")),col=text.color.false,cex=cex.samples)
text(0.675,y.max*0.2,bquote(paste(.(ICvsmu0.mat[2,2]),sep=" ")),col=text.color.false,cex=cex.samples)
}
}
if(v$cvPl == "parOri" || (v$cvPl == "parAlt" && v$mx1 == v$mx0)){
title(main=bquote(paste(mu," = ",.(v$mx1)," :",sep=" ")),col.main=color.true,adj=0,cex.main=cex.title)#"la moyenne de la population d'origine ",
y.max<-cv$maxdmx*2.1
text(0.425,y.max*0.8,bquote(paste("IC couvrent ",mu," = ",.(v$mx1)," ?",sep=" ")),cex=cex.samples)
text(0.425,y.max*0.6,bquote(paste("n",sep=" ")),cex=cex.samples)
text(0.675,y.max*0.6,bquote(paste("%",sep=" ")),cex=cex.samples)
text(0.175,y.max*0.4,"Oui",col=color.true,cex=cex.samples*0.85)
text(0.175,y.max*0.2,"Non",col=color.false,cex=cex.samples*0.85)
if(cv$n.samples>0) {
# barplot.spp<-barplot(matrix(c(100-cv$pc.ic.k.inc.allmu.vec[(v$mx1-mu.vec[1]+1)],cv$pc.ic.k.inc.allmu.vec[(v$mx1-mu.vec[1]+1)]),ncol=1),col=c(color.false,color.true), add=TRUE,beside=FALSE,space=(barplot.kH1[(v$mx1-mu.vec[1]+1)]-0.5),axes=FALSE)
ICvsmu0.mat<-matrix(c(cv$n.ic.k.inc.allmu.vec[(v$mx1-mu.vec[1]+1)],cv$n.samples-cv$n.ic.k.inc.allmu.vec[(v$mx1-mu.vec[1]+1)],cv$pc.ic.k.inc.allmu.vec[(v$mx1-mu.vec[1]+1)],100-cv$pc.ic.k.inc.allmu.vec[(v$mx1-mu.vec[1]+1)]),ncol=2)
ICvsmu0.mat<-round(ICvsmu0.mat,0)
text(0.425,y.max*0.4,bquote(paste(.(ICvsmu0.mat[1,1]),sep=" ")),col=color.true,cex=cex.samples)
text(0.675,y.max*0.4,bquote(paste(.(ICvsmu0.mat[1,2]),sep=" ")),col=color.true,cex=cex.samples)
text(0.425,y.max*0.2,bquote(paste(.(ICvsmu0.mat[2,1]),sep=" ")),col=color.false,cex=cex.samples)
text(0.675,y.max*0.2,bquote(paste(.(ICvsmu0.mat[2,2]),sep=" ")),col=color.false,cex=cex.samples)
}
}
if(v$cvPl == "parAlt" && v$mx1 != v$mx0){
title(main=bquote(paste("une valeur alternative ",mu,"'' = ",.(v$mx0)," :",sep="")),col.main=color.false,adj=0,cex.main=cex.title)
y.max<-cv$maxdmx*2.1
text(0.425,y.max*0.8,bquote(paste("IC couvrent ",mu,"'' = ",.(v$mx0)," ?",sep=" ")),cex=cex.samples)
text(0.425,y.max*0.6,bquote(paste("n",sep=" ")),cex=cex.samples)
text(0.675,y.max*0.6,bquote(paste("%",sep=" ")),cex=cex.samples)
text(0.175,y.max*0.4,"Oui",col=color.false,cex=cex.samples*0.85)
text(0.175,y.max*0.2,"Non",col=color.true,cex=cex.samples*0.85)
if(cv$n.samples>0) {
#barplot.spp<-barplot(matrix(c(100-cv$pc.ic.k.inc.allmu.vec[(v$mx0-mu.vec[1]+1)],cv$pc.ic.k.inc.allmu.vec[(v$mx0-mu.vec[1]+1)]),ncol=1),col=c(color.true,color.false), add=TRUE,beside=FALSE,space=(barplot.kH1[(v$mx0-mu.vec[1]+1)]-0.5),axes=FALSE)
ICvsmu0.mat<-matrix(c(cv$n.ic.k.inc.allmu.vec[(v$mx0-mu.vec[1]+1)],cv$n.samples-cv$n.ic.k.inc.allmu.vec[(v$mx0-mu.vec[1]+1)],cv$pc.ic.k.inc.allmu.vec[(v$mx0-mu.vec[1]+1)],100-cv$pc.ic.k.inc.allmu.vec[(v$mx0-mu.vec[1]+1)]),ncol=2)
ICvsmu0.mat<-round(ICvsmu0.mat,0)
text(0.425,y.max*0.4,bquote(paste(.(ICvsmu0.mat[1,1]),sep=" ")),col=color.false,cex=cex.samples)
text(0.675,y.max*0.4,bquote(paste(.(ICvsmu0.mat[1,2]),sep=" ")),col=color.false,cex=cex.samples)
text(0.425,y.max*0.2,bquote(paste(.(ICvsmu0.mat[2,1]),sep=" ")),col=color.true,cex=cex.samples)
text(0.675,y.max*0.2,bquote(paste(.(ICvsmu0.mat[2,2]),sep=" ")),col=color.true,cex=cex.samples)
}
}
}
}
}, height = getPlotHeight, width=getPlotWidth)
###################################################################
output$DataTable <- renderTable({
v<-getInputValues()
calc.new<-getComputedValues()
cv<-rv$cv.ls
## Transpose the sample list
if(cv$n.samples>0){
samples.as.list<-list()
for(i in 1:cv$n.samples){
samples.as.list[[i]]<-c(round(cv$samples.x.mat[i,],2),c(""),round(cv$samples.x.m.vec[i],2),round(cv$samples.x.sd.vec[i],2),c(""),round(cv$ic.k.limit.mat[i,1],2),round(cv$ic.k.limit.mat[i,2],2))
}
samples.as.matrix<- do.call(rbind,samples.as.list)
transposed.samples<-lapply(seq_len(ncol(samples.as.matrix)),function(i) samples.as.matrix[,i])
d<-data.frame(transposed.samples)
colnames(d)<-c(paste("X",1:v$n,sep="")," ","Moy","Sd"," ","LiICk","LsICk")
d
}
})
###################################################################
output$test1 <- renderText({
paste("Tab",input$Tabset,sep=" ")
})
})
0% ou .
You are about to add 0 people to the discussion. Proceed with caution.
Terminez d'abord l'édition de ce message.
Veuillez vous inscrire ou vous pour commenter