R/dynRB_VPa.R

In dynRB: Dynamic Range Boxes

dynRB_VPa <-
function(A=A, steps=201, correlogram=FALSE, row_col=c(2, 2), pca.corr=FALSE, var.thres=0.9){
  
  names(A)[1]<-"Species"
  steps0<-steps
  dims<-ncol(A)-1
  
  if(dims==1){
    
    dims<-ncol(A)-1
    insects<-levels(factor(A$Species))
    G<-expand.grid(insects,insects)
    names(G)[1:2]<-c("V1","V2")
    G$port_prod<-0
    G$port_mean<-0
    G$port_gmean<-0
    G$vol_V1_prod<-0
    G$vol_V1_mean<-0
    G$vol_V1_gmean<-0
    G$vol_V2_prod<-0
    G$vol_V2_mean<-0
    G$vol_V2_gmean<-0
    plot_data_niche_size_V1<-list()
    plot_data_niche_size_V2<-list()
    plot_data_overlap<-list()
    plot_alpha_grid<-list()
    names_data_niche_size_V1<-c()
    names_data_niche_size_V2<-c()
    for(i in 1:nrow(G)){
      S1<-subset(A,A$Species==G$V1[i])[,2:(dims+1)]
      S2<-subset(A,A$Species==G$V2[i])[,2:(dims+1)]
      SH<-A[,2:(dims+1)]
      
      V <- .volumeA2_full_dim1(S1,SH,steps=steps0,alpha_grid=seq(0,1,length=steps0) )     
      G$vol_V1_prod[i]<-V$integral_approx[1]
      G$vol_V1_mean[i]<-V$integral_approx[2]
      G$vol_V1_gmean[i]<-V$integral_approx[3]
      names_data_niche_size_V1<-c(names_data_niche_size_V1,as.character(G$V1[i]))
      plot_data_niche_size_V1[[i]]<-c(V$integral_approx[1],V$plot_data_prod)
      plot_alpha_grid[[i]]<-V$alpha_grid
      
      V <- .volumeA2_full_dim1(S2,SH,steps=steps0,alpha_grid=seq(0,1,length=steps0) )    
      G$vol_V2_prod[i]<-V$integral_approx[1]
      G$vol_V2_mean[i]<-V$integral_approx[2]
      G$vol_V2_gmean[i]<-V$integral_approx[3]
      names_data_niche_size_V2<-c(names_data_niche_size_V2,as.character(G$V2[i]))
      plot_data_niche_size_V2[[i]]<-c(V$integral_approx[1],V$plot_data_prod)
      
      EE<-.portionAinB2_full_dim1(S1,S2,steps=steps0,alpha_grid=(seq(0,1,length=steps0)[1:(steps0-1)]) )
      G$port_prod[i]<-EE$integral_approx[1]
      G$port_mean[i]<-EE$integral_approx[2]
      G$port_gmean[i]<-EE$integral_approx[3]
      plot_data_overlap[[i]]<-c(EE$integral_approx[1],EE$plot_data_prod) 
      print(i)
    }
    names(plot_data_niche_size_V1)<-names_data_niche_size_V1
    names(plot_data_niche_size_V2)<-names_data_niche_size_V2
    print(G[,c(1,2,3,6,9)])
    r<-list(plot_data_niche_size_V1=plot_data_niche_size_V1, plot_data_niche_size_V2=plot_data_niche_size_V2, plot_data_overlap=plot_data_overlap, plot_alpha_grid=plot_alpha_grid, result=G)
    invisible(r)  
  
  }else{
    
    if(pca.corr){
      A<-.trpca(A, var.thres)
    }
    
    dims<-ncol(A)-1
    
    insects<-levels(factor(A$Species))
    if(correlogram==TRUE){
      par(mfrow=row_col)
      G<-expand.grid(insects)
      h<-1
      k<-1
      for(i in 1:length(insects)){
        S1<-subset(A,A$Species==G$Var1[i])[,2:(dims+1)]
        M<-cor(S1,use="pairwise.complete.obs")
        corrplot(M, method="color",   
                 type="upper", order="hclust", addCoef.col = "black", 
                 tl.col="black", tl.srt=45,
                 sig.level = 0.01, insig = "blank", diag=FALSE, mai = c("",as.character(G$Var1[i])),mar = c(0,2,0,0))
        if(k/h==row_col[1]*row_col[2]){
          title("Correlogram for each species", outer = T)
          h<-h+1
          readline(prompt = "Press <Enter> to continue...")}
        k<-k+1
      }
      title("Correlogram for each species", outer = T)
    }
    G<-expand.grid(insects,insects)
    names(G)[1:2]<-c("V1","V2")
    G$port_prod<-0
    G$port_mean<-0
    G$port_gmean<-0
    G$vol_V1_prod<-0
    G$vol_V1_mean<-0
    G$vol_V1_gmean<-0
    G$vol_V2_prod<-0
    G$vol_V2_mean<-0
    G$vol_V2_gmean<-0
    plot_data_niche_size_V1<-list()
    plot_data_niche_size_V2<-list()
    plot_data_overlap<-list()
    plot_alpha_grid<-list()
    names_data_niche_size_V1<-c()
    names_data_niche_size_V2<-c()
    for(i in 1:nrow(G)){
      S1<-subset(A,A$Species==G$V1[i])[,2:(dims+1)]
      S2<-subset(A,A$Species==G$V2[i])[,2:(dims+1)]
      SH<-A[,2:(dims+1)]
      
      V <- .volumeA2_full(S1,SH,steps=steps0,alpha_grid=seq(0,1,length=steps0) )   
      G$vol_V1_prod[i]<-V$integral_approx[1]
      G$vol_V1_mean[i]<-V$integral_approx[2]
      G$vol_V1_gmean[i]<-V$integral_approx[3]
      names_data_niche_size_V1<-c(names_data_niche_size_V1,as.character(G$V1[i]))
      plot_data_niche_size_V1[[i]]<-c(V$integral_approx[1],V$plot_data_prod)
      plot_alpha_grid[[i]]<-V$alpha_grid
      
      V <- .volumeA2_full(S2,SH,steps=steps0,alpha_grid=seq(0,1,length=steps0) )      
      G$vol_V2_prod[i]<-V$integral_approx[1]
      G$vol_V2_mean[i]<-V$integral_approx[2]
      G$vol_V2_gmean[i]<-V$integral_approx[3]
      names_data_niche_size_V2<-c(names_data_niche_size_V2,as.character(G$V2[i]))
      plot_data_niche_size_V2[[i]]<-c(V$integral_approx[1],V$plot_data_prod)
      
      EE<-.portionAinB2_full(S1,S2,steps=steps0,alpha_grid=(seq(0,1,length=steps0)[1:(steps0-1)]) )
      G$port_prod[i]<-EE$integral_approx[1]
      G$port_mean[i]<-EE$integral_approx[2]
      G$port_gmean[i]<-EE$integral_approx[3]
      plot_data_overlap[[i]]<-c(EE$integral_approx[1],EE$plot_data_prod) 
      print(i)
    }
    names(plot_data_niche_size_V1)<-names_data_niche_size_V1
    names(plot_data_niche_size_V2)<-names_data_niche_size_V2
    print(G[,c(1,2,3,6,9)])
    r<-list(plot_data_niche_size_V1=plot_data_niche_size_V1, plot_data_niche_size_V2=plot_data_niche_size_V2, plot_data_overlap=plot_data_overlap, plot_alpha_grid=plot_alpha_grid, result=G)
    invisible(r)  
    
  }
  

}