Plot lba objects using the correspondence analysis approach as suggested by Jelihovschi (2011).
S3 methods for lba objects.
plotcorr(x, ...) ## S3 method for class 'lba.1d' plotcorr(x, xlim = NULL, ylim = NULL, xlab = NULL, ylab = NULL, metrics = TRUE, radius = rep(0.5,2), col.points = NULL, height.points = NULL, labels.points = NULL, pch.points = NULL, pos.points = NULL, args.legend = NULL, height.budget = NULL, labels.budget = NULL, pch.budget = NULL, pos.budget = NULL, cex.budget = NULL, col.budget = NULL, with.ml = c("mix","lat"), ...) ## S3 method for class 'lba.2d' plotcorr(x, dim = c(1,2), #only K = 3 xlim = NULL, ylim = NULL, xlab = NULL, ylab = NULL, args.legend = NULL, col.points = NULL, labels.points = NULL, pch.points = NULL, pos.points = NULL, labels.budget = NULL, pch.budget = NULL, pos.budget = NULL, cex.budget = NULL, col.budget = NULL, with.ml = c("mix","lat"), ...) ## S3 method for class 'lba.3d' plotcorr(x, rgl.use = FALSE, dim = c(1,2,3), #only K >= 3 xlim = NULL, ylim = NULL, zlim = NULL, xlab = NULL, ylab = NULL, zlab = NULL, args.legend = NULL, #only rgl.use=FALSE col.points = NULL, labels.points = NULL, pch.points = NULL, pos.points = NULL, labels.budget = NULL, pch.budget = NULL, pos.budget = NULL, cex.budget = NULL, col.budget = NULL, with.ml = c("mix","lat"), ...)
x: A object of lba class.dim: The dimention to be plotted. The default is c(1,2) to K = 2 and c(1,2,3) to K = 3.xlim: The x limits (x1, x2) of the plot.ylim: The y limits of the plot.zlim: The z limits of the plot.xlab: A label for the x axis, defaults to a description of "x".ylab: A label for the y axis, defaults to a description of "y".zlab: A label for the z axix, defaults to a description of "z".rgl.use: A logical value. If TRUE the 3d scatter will be done with the rgl environment, in another way the scatterplot3d will be used.metrics: Logical. If TRUE (default), the radius is plotted.radius: A arbitrary number to choose the groups. The default is 0.5. See details.col.points: The color points to be used, possibly vectors. The default is NULL. See datails.height.points: Points label height in relation to the y-coordinate.The default is NULL.labels.points: A character vector or expression specifying the text to be written. The default is NULL.pch.points: A symbols to use. O default is NULL.pos.points: A position specifier for the text. If specified this overrides any "adj" value given. Values of "1", "2", "3" and "4", respectively indicate positions below, to the left of, above and to the right of the specified coordinates.args.legend: List of additional arguments to be passed to legend; names of the list are used as argument names. Only used if K=2. The default is NULL.pch.budget: A symbols to use. O default is NULL.pos.budget: A position specifier for the text. If specified this overrides any "adj" value given. Values of "1", "2", "3" and "4", respectively indicate positions below, to the left of, above and to the right of the specified coordinates.height.budget: Budget label height in relation to the y-coordinate.The default is NULL.labels.budget: A character vector or expression specifying the text to be written. The default is NULL.cex.budget: The size of text. The default is NULL.col.budget: The color budget to be used, possibly vectors. The default is NULL.with.ml: Vector of two character strings specifying the parameters of the plot. Set "mix" to plot the mixing parameters and "lat" to plot the latent components. The default is "mix"....: Further graphical parameters.The plotcorr suggested by Jelihovschi et all (2011), has a graphical display which uses the correspondence analysis graphics of the mixing parameters and latent components matrices. In this case, a graphic display is possible for K >= 2.
The argument radius was featured in order to help the user as he or she needs do decide which are the points belonging to a certain latent budget. Only the points to the right or left of LB1 and LB2 but always towards the center of the graphic (the zero of x axis) were taken in account, since those in opposite direction automatically belong to the closest latent budget. this argument only works for K = 2. It's should be of size two.
The argument col.points takes in account the argument radius in order to color the groups which either belong or not to a certain budget, therefore, the size of the vector of this argument must be equal to the number of formed groups.
de Leeuw, J., van der Heijden, P.G.M., and Verboon, P. 1990. "A latent time budget model". Statistica Neerlandica. 44, 1, 1-21.
Jelihovschi, E.G., Alves, R.R., and Correa, F.M. 2011. Interacting latent budget analysis and correspondence analysis to analyze beauty salon management data. Biometric Brazilian Journal, 29, 657-673.
van der Ark, A. L. 1999. Contributions to Latent Budget Analysis, a tool for the analysis of comositional data. Ph.D. Thesis University of Utrecht.
Enio G. Jelihovschi (eniojelihovs@gmail.com )
Ivan Bezerra Allaman (ivanalaman@gmail.com )
plot.default, scatterplot3d, plot3d.
data('votB') K = 2 ex1 <- lba(parties ~ city, data=votB, K = 2) plotcorr(ex1) #It's very simple. with colors! plotcorr(ex1, col.points = 3:5, col.budget = c(5,3)) #Changing radius! plotcorr(ex1, radius = rep(0.7,2)) #Without metrics! plotcorr(ex1, metrics = FALSE) #Change legend options! plotcorr(ex1, args.legend = list(ncol=3)) #Change height points! plotcorr(ex1, height.points = rep(-0.1,6)) ## Not run: #K = 3 K = 3 ex2 <- lba(parties ~ city, data=votB, K = 3) plotcorr(ex2) #Change budget options plotcorr(ex2, pch.budget = 5, col.budget = 2, labels.budget = c('lba1','lba2','lba3')) #Change points options plotcorr(ex2, pch.points = 20, col.points = 4, labels.points = rownames(ex2$Aoi), args.legend = list(plot=FALSE)) #Coloring the groups plotcorr(ex2, col.points = c(1,2,2,3,3,2), col.budget = c(3,1,2), args.legend = list(ncol=3)) #K = 4 K = 4 data(postmater) new_post <- as.matrix(postmater[,-1]) row.names(new_post) <- postmater[,1] ex3 <- lba(new_post, K = K) plotcorr(ex3) #A bit didatic! plotcorr(ex3, args.legend = list(x = -2.5, y = 5.5, xpd=TRUE, ncol=5)) #Dynamic? Yes, you can! plotcorr(ex3, rgl.use = TRUE) ## End(Not run)