X-Git-Url: https://git.auder.net/?p=valse.git;a=blobdiff_plain;f=pkg%2FR%2Fplot_valse.R;h=73188d276c896c234dfe00a10149a01ab6edef65;hp=05963c8af8c6d9d598caa0aed92ebc07c9f42e70;hb=1196a43d961a95abc18d3c8e777e9a4e8233e562;hpb=4c9cc558a39c034ed75d0d5531fa0ce29d8561fc

diff --git a/pkg/R/plot_valse.R b/pkg/R/plot_valse.R
index 05963c8..73188d2 100644
--- a/pkg/R/plot_valse.R
+++ b/pkg/R/plot_valse.R
@@ -2,77 +2,86 @@
 #'
 #' It is a function which plots relevant parameters
 #'
+#' @param X matrix of covariates (of size n*p)
+#' @param Y matrix of responses (of size n*m)
+#' @param model the model constructed by valse procedure
+#' @param n sample size
+#' @param comp TRUE to enable pairwise clusters comparison
+#' @param k1 index of the first cluster to be compared
+#' @param k2 index of the second cluster to be compared
 #'
-#' @return several plots
-#'
-#' @examples TODO
+#' @importFrom ggplot2 ggplot aes ggtitle geom_tile geom_line geom_point scale_fill_gradient2 geom_boxplot theme
+#' @importFrom cowplot background_grid
+#' @importFrom reshape2 melt
 #'
 #' @export
-#'
-plot_valse = function(){
-  require("gridExtra")
-  require("ggplot2")
-  require("reshape2")
-  
+plot_valse <- function(X, Y, model, n, comp = FALSE, k1 = NA, k2 = NA)
+{
+  K <- length(model$pi)
   ## regression matrices
-  gReg = list()
-  for (r in 1:K){
-    Melt = melt(t((model$phi[,,r])))
-    gReg[[r]] = ggplot(data = Melt, aes(x=Var1, y=Var2, fill=value)) +  geom_tile() + 
-      scale_fill_gradient2(low = "blue", high = "red", mid = "white", midpoint = 0,  space = "Lab") +
-      ggtitle(paste("Regression matrices in cluster",r))
+  gReg <- list()
+  for (r in 1:K)
+  {
+    Melt <- melt(t((model$phi[, , r])))
+    gReg[[r]] <- ggplot(data = Melt, aes(x = "Var1", y = "Var2", fill = "value")) +
+      geom_tile() + scale_fill_gradient2(low = "blue", high = "red", mid = "white",
+      midpoint = 0, space = "Lab") + ggtitle(paste("Regression matrices in cluster", r))
   }
   print(gReg)
-  
+
   ## Differences between two clusters
-  k1 = 1
-  k2 = 2
-  Melt = melt(t(model$phi[,,k1]-model$phi[,,k2]))
-  gDiff = ggplot(data = Melt, aes(x=Var1, y=Var2, fill=value)) +  geom_tile() + 
-    scale_fill_gradient2(low = "blue", high = "red", mid = "white", midpoint = 0,  space = "Lab") +
-    ggtitle(paste("Difference between regression matrices in cluster",k1, "and", k2))
-  print(gDiff)
-  
-  ### Covariance matrices
-  matCov = matrix(NA, nrow = dim(model$rho[,,1])[1], ncol = K)
-  for (r in 1:K){
-    matCov[,r] = diag(model$rho[,,r])
-  }
-  MeltCov =  melt(matCov)
-  gCov = ggplot(data =MeltCov, aes(x=Var1, y=Var2, fill=value)) +  geom_tile() + 
-    scale_fill_gradient2(low = "blue", high = "red", mid = "white", midpoint = 0,  space = "Lab") +
-    ggtitle("Covariance matrices")
-  print(gCov )
-  
-  ### proportions
-  Gam = matrix(0, ncol = K, nrow = n)
-  gam  = Gam
-  for (i in 1:n){
-    for (r in 1:K){
-      sqNorm2 = sum( (Y[i,]%*%model$rho[,,r]-X[i,]%*%model$phi[,,r])^2 )
-      Gam[i,r] = model$pi[r] * exp(-0.5*sqNorm2)* det(model$rho[,,r])
-    }
-    gam[i,] = Gam[i,] / sum(Gam[i,])
-  }
-  affec = apply(gam, 1,which.max)
-  gam2 = matrix(NA, ncol = K, nrow = n)
-  for (i in 1:n){
-    gam2[i, ] = c(gam[i, affec[i]], affec[i])
+  if (comp)
+  {
+    if (is.na(k1) || is.na(k2))
+      print("k1 and k2 must be integers, representing the clusters you want to compare")
+    Melt <- melt(t(model$phi[, , k1] - model$phi[, , k2]))
+    gDiff <- ggplot(data = Melt, aes(x = "Var1", y = "Var2", fill = "value"))
+      + geom_tile()
+      + scale_fill_gradient2(low = "blue", high = "red", mid = "white", midpoint = 0,
+        space = "Lab")
+      + ggtitle(paste("Difference between regression matrices in cluster",
+        k1, "and", k2))
+    print(gDiff)
   }
-  bp <- ggplot(data.frame(gam2), aes(x=X2, y=X1, color=X2, group = X2)) +
-    geom_boxplot() + theme(legend.position = "none")
-  print(bp + background_grid(major = "xy", minor = "none"))
-  
+
+  ### Covariance matrices
+  matCov <- matrix(NA, nrow = dim(model$rho[, , 1])[1], ncol = K)
+  for (r in 1:K)
+    matCov[, r] <- diag(model$rho[, , r])
+  MeltCov <- melt(matCov)
+  gCov <- ggplot(data = MeltCov, aes(x = "Var1", y = "Var2", fill = "value")) + geom_tile()
+    + scale_fill_gradient2(low = "blue", high = "red", mid = "white", midpoint = 0,
+      space = "Lab")
+    + ggtitle("Covariance matrices")
+  print(gCov)
+
+  ### Proportions
+  gam2 <- matrix(NA, ncol = K, nrow = n)
+  for (i in 1:n)
+    gam2[i, ] <- c(model$proba[i, model$affec[i]], model$affec[i])
+
+  bp <- ggplot(data.frame(gam2), aes(x = "X2", y = "X1", color = "X2", group = "X2"))
+    + geom_boxplot()
+    + theme(legend.position = "none")
+    + background_grid(major = "xy", minor = "none")
+  print(bp)
+
   ### Mean in each cluster
-  XY = cbind(X,Y)
-  XY_class= list()
-  meanPerClass= matrix(0, ncol = K, nrow = dim(XY)[2])
-  for (r in 1:K){
-    XY_class[[r]] = XY[affec == r, ]
-    meanPerClass[,r] = apply(XY_class[[r]], 2, mean)
+  XY <- cbind(X, Y)
+  XY_class <- list()
+  meanPerClass <- matrix(0, ncol = K, nrow = dim(XY)[2])
+  for (r in 1:K)
+  {
+    XY_class[[r]] <- XY[model$affec == r, ]
+    if (sum(model$affec == r) == 1) {
+      meanPerClass[, r] <- XY_class[[r]]
+    } else {
+      meanPerClass[, r] <- apply(XY_class[[r]], 2, mean)
+    }
   }
-  data = data.frame(mean = as.vector(meanPerClass), cluster = as.character(rep(1:K, each = dim(XY)[2])), time = rep(1:dim(XY)[2],K))
-  g = ggplot(data, aes(x=time, y = mean, group = cluster, color = cluster))
-  print(g + geom_line(aes(linetype=cluster, color=cluster))+  geom_point(aes(color=cluster)) + ggtitle('Mean per cluster'))
-  
-}
\ No newline at end of file
+  data <- data.frame(mean = as.vector(meanPerClass),
+    cluster = as.character(rep(1:K, each = dim(XY)[2])), time = rep(1:dim(XY)[2], K))
+  g <- ggplot(data, aes(x = "time", y = "mean", group = "cluster", color = "cluster"))
+  print(g + geom_line(aes(linetype = "cluster", color = "cluster"))
+    + geom_point(aes(color = "cluster")) + ggtitle("Mean per cluster"))
+}