####useful function for association of variable name

finder<-function(data, vector){
  Scolnames=matrix(nrow = length(vector),ncol=ncol(data)) ;colnames(Scolnames)=colnames(data) 
  for (nvi in 1 : length(vector)){
    Scolnames[nvi,]=colnames(data)==vector[nvi]
  }
  s=apply(Scolnames,2,sum) ;sm=s*c(1:ncol(data))
  return(sm)
}


#####Parameter of transformation from datasets






pretrans<-function(data,fittab,graphpretrans=FALSE){
  
  forms=c("L.I", "L.D", "LL.TI", "LL.TD", "CC.I", "CC.D", "LC.I", "LC.D","CL.I", "CL.D", "CCC.I", "CCC.D", "CCC.TD", "CCC.TI", "LCL.I","LCL.D", "LCL.TD", "LCL.TI", "CLC.I", "CLC.D")
  tabtrans=as.data.frame(matrix(ncol=7,nrow=0))
  colnames(tabtrans)=c("variable","type","categories","s1","v1","s2","v2")    
  for (nv in 1 :nrow(fittab)){
    NVn=fittab[nv,1]
    SVN= finder(data,NVn)
    VN=data[,SVN]
    if(is.numeric(VN)==F) {
      l=levels(VN);nl=length(l)
      tabtransnv=as.data.frame(matrix(ncol=7,nrow=nl));colnames(tabtransnv)=c("variable","type","categories","s1","v1","s2","v2")    
      tabtransnv[,1]=rep(NVn,nl) 
      tabtransnv[,2]=rep("QUAL",nl)  
      tabtransnv[,3]=l
      tabtransnv[,4]=seq(from=0,to=1, length=nl)
      tabtrans=rbind(tabtrans,tabtransnv)
    } else {
      
      
      METnv= fittab[nv,2]
      if(sum(forms==METnv)==0){
        stop("The categories in the form column are not in the preset form defined in the function pretrans")
      }
      if(METnv=="L.I"){
        tabtransnv=as.data.frame(matrix(ncol=7,nrow=1));colnames(tabtransnv)=c("variable","type","categories","s1","v1","s2","v2")    
        tabtransnv[,1]=NVn
        tabtransnv[,2]="QUAN"
       
        tabtransnv[,4]=0 
        tabtransnv[,5]=min(VN,na.rm =TRUE)
        tabtransnv[,6]=1
        tabtransnv[,7]=max(VN,na.rm =TRUE)
        tabtrans=rbind(tabtrans,tabtransnv)
      }
      if(METnv=="L.D"){
        tabtransnv=as.data.frame(matrix(ncol=7,nrow=1));colnames(tabtransnv)=c("variable","type","categories","s1","v1","s2","v2")    
        tabtransnv[,1]=NVn
        tabtransnv[,2]="QUAN"
        tabtransnv[,4]=1 
        tabtransnv[,5]=min(VN,na.rm =TRUE)
        tabtransnv[,6]=0
        tabtransnv[,7]=max(VN,na.rm =TRUE)
        tabtrans=rbind(tabtrans,tabtransnv)
      }
      if(METnv=="LL.TI"){
        
        S1=((max(VN,na.rm =TRUE)-min(VN,na.rm =TRUE))/2)+min(VN,na.rm =TRUE)
        
        tabtransnv=as.data.frame(matrix(ncol=7,nrow=2));colnames(tabtransnv)=c("variable","type","categories","s1","v1","s2","v2")    
        tabtransnv[,1]=rep(NVn,2)
        tabtransnv[,2]="QUAN"
        tabtransnv[,4]=c(0,1) 
        tabtransnv[,5]=c(min(VN,na.rm =TRUE),S1)
        tabtransnv[,6]=c(1,0)
        tabtransnv[,7]=c(S1,max(VN,na.rm =TRUE))
        tabtrans=rbind(tabtrans,tabtransnv)
      }
      if(METnv=="LL.TD"){
        
        S1=((max(VN,na.rm =TRUE)-min(VN,na.rm =TRUE))/2)+min(VN,na.rm =TRUE)
        
        tabtransnv=as.data.frame(matrix(ncol=7,nrow=2));colnames(tabtransnv)=c("variable","type","categories","s1","v1","s2","v2")    
        tabtransnv[,1]=rep(NVn,2)
        tabtransnv[,2]="QUAN"
       
        tabtransnv[,4]=c(1,0) 
        tabtransnv[,5]=c(min(VN,na.rm =TRUE),S1)
        tabtransnv[,6]=c(0,1)
        tabtransnv[,7]=c(S1,max(VN,na.rm =TRUE))
        tabtrans=rbind(tabtrans,tabtransnv)
      }
      if(METnv=="CC.I"){
        
        S1=((max(VN,na.rm =TRUE)-min(VN,na.rm =TRUE))/2)+min(VN,na.rm =TRUE)
        
        tabtransnv=as.data.frame(matrix(ncol=7,nrow=2));colnames(tabtransnv)=c("variable","type","categories","s1","v1","s2","v2")    
        tabtransnv[,1]=rep(NVn,2)
        tabtransnv[,2]="QUAN"
        tabtransnv[,4]=c(0,1)
        tabtransnv[,5]=c(min(VN,na.rm =TRUE),S1)
        tabtransnv[,6]=c(0,1) 
        tabtransnv[,7]=c(S1,max(VN,na.rm =TRUE))
        
        tabtrans=rbind(tabtrans,tabtransnv)
      }
      
      if(METnv=="CC.D"){
        
        S1=((max(VN,na.rm =TRUE)-min(VN,na.rm =TRUE))/2)+min(VN,na.rm =TRUE)
        
        tabtransnv=as.data.frame(matrix(ncol=7,nrow=2));colnames(tabtransnv)=c("variable","type","categories","s1","v1","s2","v2")    
        tabtransnv[,1]=rep(NVn,2)
        tabtransnv[,2]="QUAN"
        tabtransnv[,4]=c(1,0) 
        tabtransnv[,5]=c(min(VN,na.rm =TRUE),S1)
        tabtransnv[,6]=c(1,0)
        tabtransnv[,7]=c(S1,max(VN,na.rm =TRUE))
             
        
        tabtrans=rbind(tabtrans,tabtransnv)
      } 
      if(METnv=="LC.I"){
        
        S1=((max(VN,na.rm =TRUE)-min(VN,na.rm =TRUE))/2)+min(VN,na.rm =TRUE)
        
        tabtransnv=as.data.frame(matrix(ncol=7,nrow=2));colnames(tabtransnv)=c("variable","type","categories","s1","v1","s2","v2")    
        tabtransnv[,1]=rep(NVn,2)
        tabtransnv[,2]="QUAN"
        tabtransnv[,4]=c(0,1) 
        tabtransnv[,5]=c(min(VN,na.rm =TRUE),S1)
        tabtransnv[,6]=c(1,1)
        tabtransnv[,7]=c(S1,max(VN,na.rm =TRUE))
        tabtrans=rbind(tabtrans,tabtransnv)
      } 
      if(METnv=="LC.D"){
        
        S1=((max(VN,na.rm =TRUE)-min(VN,na.rm =TRUE))/2)+min(VN,na.rm =TRUE)
        
        tabtransnv=as.data.frame(matrix(ncol=7,nrow=2));colnames(tabtransnv)=c("variable","type","categories","s1","v1","s2","v2")    
        tabtransnv[,1]=rep(NVn,2)
        tabtransnv[,2]="QUAN"
        tabtransnv[,4]=c(1,0)
        tabtransnv[,5]=c(min(VN,na.rm =TRUE),S1)
        tabtransnv[,6]=c(0,0)
        tabtransnv[,7]=c(S1,max(VN,na.rm =TRUE))
        
        tabtrans=rbind(tabtrans,tabtransnv)
      } 
      if(METnv=="CL.I"){
        
        S1=((max(VN,na.rm =TRUE)-min(VN,na.rm =TRUE))/2)+min(VN,na.rm =TRUE)
        
        tabtransnv=as.data.frame(matrix(ncol=7,nrow=2));colnames(tabtransnv)=c("variable","type","categories","s1","v1","s2","v2")    
        tabtransnv[,1]=rep(NVn,2)
        tabtransnv[,2]="QUAN"
        tabtransnv[,4]=c(0,0)
        tabtransnv[,5]=c(min(VN,na.rm =TRUE),S1)
        tabtransnv[,6]=c(0,1)
        tabtransnv[,7]=c(S1,max(VN,na.rm =TRUE))
        
        tabtrans=rbind(tabtrans,tabtransnv)
      } 
      if(METnv=="CL.D"){
        
        S1=((max(VN,na.rm =TRUE)-min(VN,na.rm =TRUE))/2)+min(VN,na.rm =TRUE)
        
        tabtransnv=as.data.frame(matrix(ncol=7,nrow=2));colnames(tabtransnv)=c("variable","type","categories","s1","v1","s2","v2")    
        tabtransnv[,1]=rep(NVn,2)
        tabtransnv[,2]="QUAN"
        tabtransnv[,4]=c(1,1)
        tabtransnv[,5]=c(min(VN,na.rm =TRUE),S1)
        tabtransnv[,6]=c(1,0)
        tabtransnv[,7]=c(S1,max(VN,na.rm =TRUE))
       
        tabtrans=rbind(tabtrans,tabtransnv)
      }
      if(METnv=="CCC.I"){
        
        S1=((max(VN,na.rm =TRUE)-min(VN,na.rm =TRUE))/3)+min(VN,na.rm =TRUE)
        S2=(2*(max(VN,na.rm =TRUE)-min(VN,na.rm =TRUE))/3)+min(VN,na.rm =TRUE)
        
        tabtransnv=as.data.frame(matrix(ncol=7,nrow=3));colnames(tabtransnv)=c("variable","type","categories","s1","v1","s2","v2")    
        tabtransnv[,1]=rep(NVn,3)
        tabtransnv[,2]="QUAN"
        tabtransnv[,4]=c(0,0.5,1) 
        tabtransnv[,6]=c(0,0.5,1)
        tabtransnv[,5]=c(min(VN,na.rm =TRUE),S1,S2)
        tabtransnv[,7]=c(S1,S2,max(VN,na.rm =TRUE))
     
        tabtrans=rbind(tabtrans,tabtransnv)
      }
      if(METnv=="CCC.D"){
        S1=((max(VN,na.rm =TRUE)-min(VN,na.rm =TRUE))/3)+min(VN,na.rm =TRUE)
        S2=(2*(max(VN,na.rm =TRUE)-min(VN,na.rm =TRUE))/3)+min(VN,na.rm =TRUE)
        
        tabtransnv=as.data.frame(matrix(ncol=7,nrow=3));colnames(tabtransnv)=c("variable","type","categories","s1","v1","s2","v2")    
        tabtransnv[,1]=rep(NVn,3)
        tabtransnv[,2]="QUAN"
        tabtransnv[,4]=c(1,0.5,0) 
        tabtransnv[,5]=c(min(VN,na.rm =TRUE),S1,S2)
        tabtransnv[,6]=c(1,0.5,0)
        tabtransnv[,7]=c(S1,S2,max(VN,na.rm =TRUE))
        
        tabtrans=rbind(tabtrans,tabtransnv)
      }
      if(METnv=="CCC.TI"){
        S1=((max(VN,na.rm =TRUE)-min(VN,na.rm =TRUE))/3)+min(VN,na.rm =TRUE)
        S2=(2*(max(VN,na.rm =TRUE)-min(VN,na.rm =TRUE))/3)+min(VN,na.rm =TRUE)
        
        tabtransnv=as.data.frame(matrix(ncol=7,nrow=3));colnames(tabtransnv)=c("variable","type","categories","s1","v1","s2","v2")    
        tabtransnv[,1]=rep(NVn,3)
        tabtransnv[,2]="QUAN"
        tabtransnv[,4]=c(0,1,0)
        tabtransnv[,5]=c(min(VN,na.rm =TRUE),S1,S2)
        tabtransnv[,6]=c(0,1,0) 
        tabtransnv[,7]=c(S1,S2,max(VN,na.rm =TRUE))
       
        tabtrans=rbind(tabtrans,tabtransnv)
      }
      if(METnv=="CCC.TD"){
        S1=((max(VN,na.rm =TRUE)-min(VN,na.rm =TRUE))/3)+min(VN,na.rm =TRUE)
        S2=(2*(max(VN,na.rm =TRUE)-min(VN,na.rm =TRUE))/3)+min(VN,na.rm =TRUE)
        
        tabtransnv=as.data.frame(matrix(ncol=7,nrow=3));colnames(tabtransnv)=c("variable","type","categories","s1","v1","s2","v2")    
        tabtransnv[,1]=rep(NVn,3)
        tabtransnv[,2]="QUAN"
        tabtransnv[,4]=c(1,0,1)
        tabtransnv[,5]=c(min(VN,na.rm =TRUE),S1,S2)
        tabtransnv[,6]=c(1,0,1)
        tabtransnv[,7]=c(S1,S2,max(VN,na.rm =TRUE))
   
        tabtrans=rbind(tabtrans,tabtransnv)
      }
      
      if(METnv=="LCL.I"){
        S1=((max(VN,na.rm =TRUE)-min(VN,na.rm =TRUE))/3)+min(VN,na.rm =TRUE)
        S2=(2*(max(VN,na.rm =TRUE)-min(VN,na.rm =TRUE))/3)+min(VN,na.rm =TRUE)
        tabtransnv=as.data.frame(matrix(ncol=7,nrow=3));colnames(tabtransnv)=c("variable","type","categories","s1","v1","s2","v2")    
        tabtransnv[,1]=rep(NVn,3)
        tabtransnv[,2]="QUAN"
        tabtransnv[,4]=c(0,0.5,0.5) 
        tabtransnv[,5]=c(min(VN,na.rm =TRUE),S1,S2)
        tabtransnv[,6]=c(0.5,0.5,1)
        tabtransnv[,7]=c(S1,S2,max(VN,na.rm =TRUE))
        
        tabtrans=rbind(tabtrans,tabtransnv)
        
      }
      
      if(METnv=="LCL.D"){
        S1=((max(VN,na.rm =TRUE)-min(VN,na.rm =TRUE))/3)+min(VN,na.rm =TRUE)
        S2=(2*(max(VN,na.rm =TRUE)-min(VN,na.rm =TRUE))/3)+min(VN,na.rm =TRUE)
        
        tabtransnv=as.data.frame(matrix(ncol=7,nrow=3));colnames(tabtransnv)=c("variable","type","categories","s1","v1","s2","v2")    
        tabtransnv[,1]=rep(NVn,3)
        tabtransnv[,2]="QUAN"
        tabtransnv[,4]=c(1,0.5,0.5)
        tabtransnv[,5]=c(min(VN,na.rm =TRUE),S1,S2)
        tabtransnv[,6]=c(0.5,0.5,0)
        tabtransnv[,7]=c(S1,S2,max(VN,na.rm =TRUE))
        
        tabtrans=rbind(tabtrans,tabtransnv)
        
      }
      
      
      if(METnv=="LCL.TI"){
        S1=((max(VN,na.rm =TRUE)-min(VN,na.rm =TRUE))/3)+min(VN,na.rm =TRUE)
        S2=(2*(max(VN,na.rm =TRUE)-min(VN,na.rm =TRUE))/3)+min(VN,na.rm =TRUE)
        tabtransnv=as.data.frame(matrix(ncol=7,nrow=3));colnames(tabtransnv)=c("variable","type","categories","s1","v1","s2","v2")    
        tabtransnv[,1]=rep(NVn,3)
        tabtransnv[,2]="QUAN"
        tabtransnv[,4]=c(0,1,1) 
        tabtransnv[,5]=c(min(VN,na.rm =TRUE),S1,S2)
        tabtransnv[,6]=c(1,1,0)
        tabtransnv[,7]=c(S1,S2,max(VN,na.rm =TRUE))
        
        tabtrans=rbind(tabtrans,tabtransnv)
        
      }
      
      if(METnv=="LCL.TD"){
        S1=((max(VN,na.rm =TRUE)-min(VN,na.rm =TRUE))/3)+min(VN,na.rm =TRUE)
        S2=(2*(max(VN,na.rm =TRUE)-min(VN,na.rm =TRUE))/3)+min(VN,na.rm =TRUE)
        tabtransnv=as.data.frame(matrix(ncol=7,nrow=3));colnames(tabtransnv)=c("variable","type","categories","s1","v1","s2","v2")    
        tabtransnv[,1]=rep(NVn,3)
        tabtransnv[,2]="QUAN"
        tabtransnv[,4]=c(1,0,0) 
        tabtransnv[,5]=c(min(VN,na.rm =TRUE),S1,S2)
        tabtransnv[,6]=c(0,0,1)
        tabtransnv[,7]=c(S1,S2,max(VN,na.rm =TRUE))
        
        tabtrans=rbind(tabtrans,tabtransnv)
        
      }
      
      if(METnv=="CLC.I"){
        S1=((max(VN,na.rm =TRUE)-min(VN,na.rm =TRUE))/3)+min(VN,na.rm =TRUE)
        S2=(2*(max(VN,na.rm =TRUE)-min(VN,na.rm =TRUE))/3)+min(VN,na.rm =TRUE)
        tabtransnv=as.data.frame(matrix(ncol=7,nrow=3));colnames(tabtransnv)=c("variable","type","s1","v1","s2","v2")    
        tabtransnv[,1]=rep(NVn,3)
        tabtransnv[,2]="QUAN"
        tabtransnv[,4]=c(0,0,1)
        tabtransnv[,5]=c(min(VN,na.rm =TRUE),S1,S2)
        tabtransnv[,6]=c(0,1,1)
        tabtransnv[,7]=c(S1,S2,max(VN,na.rm =TRUE))
        
        tabtrans=rbind(tabtrans,tabtransnv)
        
      }
      if(METnv=="CLC.D"){
        S1=((max(VN,na.rm =TRUE)-min(VN,na.rm =TRUE))/3)+min(VN,na.rm =TRUE)
        S2=(2*(max(VN,na.rm =TRUE)-min(VN,na.rm =TRUE))/3)+min(VN,na.rm =TRUE)
        tabtransnv=as.data.frame(matrix(ncol=7,nrow=3));colnames(tabtransnv)=c("variable","type","categories","s1","v1","s2","v2")    
        tabtransnv[,1]=rep(NVn,3)
        tabtransnv[,2]="QUAN"
        tabtransnv[,4]=c(1,1,0)
        tabtransnv[,5]=c(min(VN,na.rm =TRUE),S1,S2)
        tabtransnv[,6]=c(1,0,0)
        tabtransnv[,7]=c(S1,S2,max(VN,na.rm =TRUE)) 
        
        tabtrans=rbind(tabtrans,tabtransnv)
        
      }
      
    }
    
    
    
  }
  if(graphpretrans==TRUE){
  
  datat=trans(data=data,tabtrans = tabtrans)
  ng=ncol(datat)
  nc=ceiling(sqrt(ng))
  nr=floor(sqrt(ng))
  par(mfrow=c(nc,nr))
  par(mar=c(3,3,0.5,0.5)) 
  
  for(ngi in 1: ng){
    
    Vti=colnames(datat)[ngi]
    
    plot(datat[,Vti]~data[,Vti],ylim=c(-0.1,1.1),ylab="Notation",main=Vti,pch=19,yaxt="n",bty="l")
    axis(2,at=c(0,0.5,1))
    
    
  }
  }
  return(tabtrans)
}   




######function of transformation at individuals scale
transf<-function(var,type,S1,V1,S2,V2){
  if(type=="QUAL"){
    S=S1
  }
  if(type=="QUAN"){
    
    a=(S2-S1)/(V2-V1)
    b=S1-(a*V1)
    
    S=a*var+b
  }
  
  return(S)
}

######function of aggregation at individuals scale
aggrind<-function(V,MET,P){
  
  if(MET=="MEAN"){
    O=weighted.mean(V,P,na.rm=TRUE)
  }
  if(MET=="MIN"){
    O=min(V,na.rm=TRUE)
  }
  if(MET=="MAX"){
    O=max(V,na.rm=TRUE)
  }
  
  return(O)
}

##############function of transformation for datasets
trans<-function(data,tabtrans){
  level=levels(tabtrans$variable)
  datat=matrix(nrow=nrow(data),ncol=length(level));colnames(datat)=level;rownames(datat)=rownames(data)
  
  
  for( y in 1:ncol(datat)){
    Vy=colnames(datat)[y]
    Sy=finder(data,Vy) 
    if(sum(Sy)==0){
      stop("The variable in the table of transformation are not in the data")
    }
    v=data[,Sy]
    tabV=subset(tabtrans,tabtrans$variable==Vy)
    vt=vector()
    
    for (nr in 1 : length(v)){
      vnr=v[nr]
      if(is.na(vnr)==TRUE){
        vt[nr]=NA 
      }
      if(is.na(vnr)==FALSE){
        if(tabV$type[1]=="QUAN"){
          for(ns in 1 : nrow(tabV)){
            SBns=tabV$v1[ns];SHns=tabV$v2[ns]
            if(SBns>= SHns){
              stop("The value v1 must be lower than the value v2 in the table")  
            }
            if(vnr>=SBns&vnr<=SHns){ 
              vt[nr]=transf(vnr,type="QUAN",S1=tabV$s1[ns],V1=tabV$v1[ns],S2=tabV$s2[ns],V2=tabV$v2[ns])
            }
          }
        }
        if(tabV$type[1]=="QUAL"){
          
          for(ns in 1 : nrow(tabV)){
            T=tabV$categories[ns]
            if(as.character(vnr)==as.character(T)){
              vt[nr]=transf(vnr,type="QUAL",S1=tabV$s1[ns])
            }
          }
        }
        
      }
    }
    
    Vy=colnames(datat)[y]
    Syt=finder(datat,Vy) 
    datat[,Syt]=vt
    
  }
  
  
  return(datat)
  
}



##############function of aggregation for datasets
aggr<-function(datat,tabaggr){
  if(max(datat,na.rm=T)>1|min(datat.na.rm=T)<0){
    stop("The notation after the transformation are not between 0 and 1 ")
  }
  tabaggr$ON <- factor(tabaggr$ON)
  OO=tabaggr$ON[1]
  for (i in 2: nrow(tabaggr)){
    if(tabaggr$ON[i-1]!=tabaggr$ON[i]){
      
      OO=c(OO,tabaggr$ON[i]) 
    }
  }
  
  nO=levels(tabaggr$ON)[OO]
  resO=matrix(ncol=length(nO),nrow=nrow(datat))
  colnames(resO)=nO;rownames(resO)=rownames(datat)
  restot=cbind(datat,resO)
  
  
  for (nbO in 1: length(nO)){
    nOi=nO[nbO]
    tagr=subset(tabaggr,tabaggr$ON==nOi)
     vecvnr=tagr$IN
      svnr=finder(restot,vecvnr)
      if(max(svnr)>ncol(restot)){
                stop("The variable in the IN section are not wll writen")
      }
      for(vi in 1: length(vecvnr)){
        vni=vecvnr[vi]
        svi=finder(restot,vni)
            if(sum(svi)==0){
        stop("The variable in the IN column in the table of aggregation are not in the data or in the ON column of the table")
      }
    }
    for (nr in 1: nrow(restot)){
     
      vnr=restot[nr,svnr]
      M=tagr$method[1]
      P=tagr$weight
      restot[nr,nOi]=aggrind(V=vnr,MET=M,P=P)
      
    }
  }
  
  if(is.na(sum(datat))==T){
    warning("missing data  are presented in the input notation after the transformation; in the aggregation, the missing values are not considered and the aggregation are made with the other variables ",call.=FALSE)
  }
  resaggr=list()
  sno=nO;fsno=finder(restot,sno)
  resaggr=restot[,fsno]
  return(resaggr)
  
  
  
}

#####function TATALE
tatale<-function(data,tabtrans,tabaggr,graphtrans=FALSE,graphaggr=FALSE){
  datat=trans(data=data,tabtrans = tabtrans)
  resaggr=aggr(datat,tabaggr)
  if(graphtrans==T){
    ng=ncol(datat)
    nc=ceiling(sqrt(ng))
    nr=floor(sqrt(ng))
    par(mfrow=c(nc,nr))
    par(mar=c(3,3,0.5,0.5)) 
    
    for(ngi in 1: ng){
    
      Vti=colnames(datat)[ngi]
    
      plot(datat[,Vti]~data[,Vti],ylim=c(-0.1,1.1),ylab="Scores",main=Vti,pch=19,yaxt="n",bty="l")
      axis(2,at=c(0,0.5,1))
    }
  }
  if(graphaggr==TRUE){
    plot.new()
    par(mfrow=c(1,1))
    graphaggr(tabaggr,orientation="LR",space=1)
  }
  
res=list()
res$aggr=resaggr
res$trans=datat
  
  return(res)
}
###input
graphaggr=function(tabaggr,orientation="LR",space=1){
  
  number=vector(length=nrow(tabaggr))
  ni=1
  
  for (ini in 1: nrow(tabaggr)){
    inv=tabaggr[ini,2]  
    s=sum(tabaggr[,1]==as.character(inv[1]))
    number[ini]=ifelse(s==0,ni,"X")
  }
  n=1:nrow(tabaggr)
  while(sum(number=="X")>0){
    ni=ni+1
    nni=subset(n,number=="X")
    
    tabni=subset(tabaggr,number=="X")
    numberi=vector(length=(nrow(tabni)))
    for (ini in 1: nrow(tabni)){
      inv=tabni[ini,2]  
      s=sum(tabni[,1]==as.character(inv[1]))
      numberi[ini]=ifelse(s==0,ni,"X")
    }
    
    number[nni]=numberi
    
  }
  
  numberin=as.data.frame(as.numeric(number))
  cori=cbind(tabaggr$IN,numberin);colnames(cori)=c("V","L")
  
  
  #output
  ni=ni+1
  numbero=vector()
  for (ini in 1: nrow(tabaggr)){
    inv=tabaggr[ini,1]
    s=sum(tabaggr[,2]==as.character(inv[1]))
    numbero[ini]=ifelse(s==0,ni,"X")
  }
  tabao=subset(tabaggr,numbero!="X") 
  lvo=levels(as.factor(as.character(tabao$ON)))
  
  nio=rep(ni,length(lvo))
  coro=cbind(lvo,nio);colnames(coro)=c("V","L")
  cor=rbind(cori,coro)
  cord=matrix(ncol=3,nrow=0);colnames(cord)=c("V","L","y")
  for(lni in 1:max(cor$L)){
    corli=subset(cor,cor$L==lni)
    yli=nrow(corli)/2
    ei=lni*space+1-space
    sli=seq(from=yli,to=-yli)*ei;sli=sli[1:nrow(corli)];
    corlii=cbind(corli,sli);colnames(corlii)=colnames(cord);cord=rbind(cord,corlii)
  }
  
  if (orientation=="LR"){
    lmax=as.numeric(max(cord$L))+1
    plot(cord$y~cord$L,pch=19,xlim=c(0,lmax),ylab="",xlab="",yaxt="n",xaxt="n",cex=0.8)
    
    for (lki in 1: nrow(tabaggr)){
      iki=as.character(tabaggr[lki,2]);corki=cord[cord$V==iki,]  
      oki=as.character(tabaggr[lki,1]);corko=cord[cord$V==oki,] 
      
      ly=ifelse(tabaggr[lki,3]=="MEAN",1,ifelse(tabaggr[lki,3]=="MIN",2,3))
      cork=rbind(corki,corko)
      lines(cork$y~cork$L,lty=ly)
      if(tabaggr[lki,3]=="MEAN"){
        text(corki$L,corki$y,labels=tabaggr[lki,4],adj=c(-0.8,-0.8),cex=0.6)
      }
    }
    
    cord1=subset(cord,cord$L==1)  
    text(cord1$L,cord1$y,labels=cord1$V,adj=1.1,col=1)
    
    cordr1=subset(cord,cord$L!=1)  
    text(cordr1$L,cordr1$y,labels=cordr1$V,adj=c(-0.2,0.6),col=grey(0.5))
    legend("bottomright",lty=c(1,2,3),legend=c("MEAN","MIN","MAX"),cex=0.8)
  }
  if (orientation=="TD"){
    
    cord$L=-as.numeric(cord$L)
    lmax=as.numeric(min(cord$L))-1
    
    plot(cord$L~cord$y,pch=19,ylim=c(lmax,0),ylab="",xlab="",yaxt="n",xaxt="n",cex=0.8)
    for (lki in 1: nrow(tabaggr)){
      iki=as.character(tabaggr[lki,2]);corki=cord[cord$V==iki,]  
      oki=as.character(tabaggr[lki,1]);corko=cord[cord$V==oki,] 
      
      ly=ifelse(tabaggr[lki,3]=="MEAN",1,ifelse(tabaggr[lki,3]=="MIN",2,3))
      cork=rbind(corki,corko)
      lines(cork$L~cork$y,lty=ly)
      if(tabaggr[lki,3]=="MEAN"){
        text(corki$y+0.1,corki$L-0.2,labels=tabaggr[lki,4],adj=c(0,1),cex=0.6)
      }
    }
    cord1=subset(cord,cord$L==-1)  
    text(cord1$y,cord1$L,labels=cord1$V,adj=c(0,0.5),col=1,srt=90,cex=0.8,offset=1)
    
    cordr1=subset(cord,cord$L!=-1)  
    text(cordr1$y,cordr1$L,labels=cordr1$V,adj=c(1,1),col=grey(0.5))
    legend("bottomright",lty=c(1,2,3),legend=c("MEAN","MIN","MAX"),cex=0.8)  
  }
}



graphtrans= function(tabtrans){
  
  lv=levels(tabtrans$variable)
  s=split(tabtrans,tabtrans$variable)
  datam=as.data.frame(matrix(nrow=100,ncol=length(lv)));colnames(datam)=lv
  for( sni in 1: length(lv)){
    tabs=s[[sni]]
    if(tabs[1,2]=="QUAN"){
      minn=min(tabs$v1)
      maxn=max(tabs$v2)
      datam[,sni]=seq(from=minn,to=maxn, length.out = 100)
    }
    if(tabs[1,2]=="QUAL"){
      nc=levels(tabs$categories)
      datam[,sni]=as.factor(rep(tabs$categories,length.out=100))
    } 
  }
  datat=trans(data=datam,tabtrans = tabtrans)
  ng=ncol(datat)
  nc=ceiling(sqrt(ng))
  nr=floor(sqrt(ng))
  par(mfrow=c(nc,nr))
  par(mar=c(3,3,0.5,0.5)) 
  
  for(ngi in 1: ng){
    
    Vti=colnames(datat)[ngi]
    
    plot(datat[,Vti]~datam[,Vti],ylim=c(-0.1,1.1),ylab="Scores",main=Vti,pch=19,yaxt="n",bty="l")
    axis(2,at=c(0,0.5,1))
    
  }
    
}