| 1 | #' Plot |
| 2 | #' |
| 3 | #' It is a function which plots relevant parameters |
| 4 | #' |
| 5 | #' |
| 6 | #' @return several plots |
| 7 | #' |
| 8 | #' @examples TODO |
| 9 | #' |
| 10 | #' @export |
| 11 | #' |
| 12 | plot_valse = function(){ |
| 13 | require("gridExtra") |
| 14 | require("ggplot2") |
| 15 | require("reshape2") |
| 16 | |
| 17 | ## regression matrices |
| 18 | gReg = list() |
| 19 | for (r in 1:K){ |
| 20 | Melt = melt(t((model$phi[,,r]))) |
| 21 | gReg[[r]] = ggplot(data = Melt, aes(x=Var1, y=Var2, fill=value)) + geom_tile() + |
| 22 | scale_fill_gradient2(low = "blue", high = "red", mid = "white", midpoint = 0, space = "Lab") + |
| 23 | ggtitle(paste("Regression matrices in cluster",r)) |
| 24 | } |
| 25 | print(gReg) |
| 26 | |
| 27 | ## Differences between two clusters |
| 28 | k1 = 1 |
| 29 | k2 = 2 |
| 30 | Melt = melt(t(model$phi[,,k1]-model$phi[,,k2])) |
| 31 | gDiff = ggplot(data = Melt, aes(x=Var1, y=Var2, fill=value)) + geom_tile() + |
| 32 | scale_fill_gradient2(low = "blue", high = "red", mid = "white", midpoint = 0, space = "Lab") + |
| 33 | ggtitle(paste("Difference between regression matrices in cluster",k1, "and", k2)) |
| 34 | print(gDiff) |
| 35 | |
| 36 | ### Covariance matrices |
| 37 | matCov = matrix(NA, nrow = dim(model$rho[,,1])[1], ncol = K) |
| 38 | for (r in 1:K){ |
| 39 | matCov[,r] = diag(model$rho[,,r]) |
| 40 | } |
| 41 | MeltCov = melt(matCov) |
| 42 | gCov = ggplot(data =MeltCov, aes(x=Var1, y=Var2, fill=value)) + geom_tile() + |
| 43 | scale_fill_gradient2(low = "blue", high = "red", mid = "white", midpoint = 0, space = "Lab") + |
| 44 | ggtitle("Covariance matrices") |
| 45 | print(gCov ) |
| 46 | |
| 47 | ### proportions |
| 48 | Gam = matrix(0, ncol = K, nrow = n) |
| 49 | gam = Gam |
| 50 | for (i in 1:n){ |
| 51 | for (r in 1:K){ |
| 52 | sqNorm2 = sum( (Y[i,]%*%model$rho[,,r]-X[i,]%*%model$phi[,,r])^2 ) |
| 53 | Gam[i,r] = model$pi[r] * exp(-0.5*sqNorm2)* det(model$rho[,,r]) |
| 54 | } |
| 55 | gam[i,] = Gam[i,] / sum(Gam[i,]) |
| 56 | } |
| 57 | affec = apply(gam, 1,which.max) |
| 58 | gam2 = matrix(NA, ncol = K, nrow = n) |
| 59 | for (i in 1:n){ |
| 60 | gam2[i, ] = c(gam[i, affec[i]], affec[i]) |
| 61 | } |
| 62 | bp <- ggplot(data.frame(gam2), aes(x=X2, y=X1, color=X2, group = X2)) + |
| 63 | geom_boxplot() + theme(legend.position = "none") |
| 64 | print(bp + background_grid(major = "xy", minor = "none")) |
| 65 | |
| 66 | ### Mean in each cluster |
| 67 | XY = cbind(X,Y) |
| 68 | XY_class= list() |
| 69 | meanPerClass= matrix(0, ncol = K, nrow = dim(XY)[2]) |
| 70 | for (r in 1:K){ |
| 71 | XY_class[[r]] = XY[affec == r, ] |
| 72 | meanPerClass[,r] = apply(XY_class[[r]], 2, mean) |
| 73 | } |
| 74 | data = data.frame(mean = as.vector(meanPerClass), cluster = as.character(rep(1:K, each = dim(XY)[2])), time = rep(1:dim(XY)[2],K)) |
| 75 | g = ggplot(data, aes(x=time, y = mean, group = cluster, color = cluster)) |
| 76 | print(g + geom_line(aes(linetype=cluster, color=cluster))+ geom_point(aes(color=cluster)) + ggtitle('Mean per cluster')) |
| 77 | |
| 78 | } |