R/Plots.R
In mariytu/RegressionLibs: What the package does (short line)
Defines functions
elbowPlot
ScatterplotMatrix
ParallelPlot
DensityPlot
Plot3D
simplePlot
RidgePlot
Documented in
DensityPlot
elbowPlot
ParallelPlot
Plot3D
ScatterplotMatrix
simplePlot
#' Elbow Plot for PCA (Plot)
#'
#' Generate a plot of 10 first variances of Principal Components. This is useful to
#' determinate which are the most important components.
#'
#' @param data.pca a list with class "prcomp" containing all principal components
#' calculated.
#' @seealso CalculateVariance, plotPC
#' @examples
#' #Example 1
#' iris.x <- iris[,1:4] # These are the independent variables
#' # We know that there are no missing values in the data set
#'
#' # performing prcomp
#' ir.pca <- prcomp(iris.x, center = TRUE, scale. = TRUE)
#'
#' # Generating elbow plot to detect the most important principal components
#' elbowPlot(ir.pca)
#'
#'
#' #Example 2
#' # Getting a clean data set (without missing values)
#' cars <- read.csv("https://dl.dropboxusercontent.com/u/12599702/autosclean.csv", sep = ";", dec = ",")
#' cars.x <- cars[,1:16] # These are the independent variables
#'
#' # Performing prcomp
#' cars.pca <- prcomp(cars.x, center = TRUE, scale. = TRUE)
#'
#' # Generating elbow plot to detect the most important principal components
#' elbowPlot(cars.pca)
elbowPlot <- function(data.pca) {
if (missing(data.pca)) {
stop("Need to specify data.pca!")
}
if (class(data.pca) != "prcomp") {
stop("data.pca must be a prcomp class!")
}
#All parameters are OK!
rowsData <- length(data.pca$sdev)
seqRow <- seq(from = 1, to = rowsData, length.out = rowsData)
dataPlot <- data.frame(seqRow, data.pca$sdev)
names(dataPlot) <- c("PCA", "Variances")
if (nrow(dataPlot)>10) {
dataPlot <- dataPlot[1:10,]
}
dataPlot <- CalculateVariance(dataPlot, 2)
p <- ggplot(data = dataPlot, aes(x = PCA, y = Variances, group = 1)) +
geom_line(colour = "dodgerblue4", alpha = 0.5, size = 1) +
geom_point(colour = "dodgerblue4", size = 2, alpha = 0.5) +
expand_limits(y = 0) +
xlab("PCs") + ylab("Variances") +
scale_x_continuous(breaks = dataPlot$PCA) +
theme(panel.grid.minor = element_blank(), #remove gridlines
legend.position = "bottom" #legend at the bottom
)#end theme
return (p)
}
#' Scatterplot Matrix (Plot)
#'
#' Generate a Scatterplot Matrix of some columns of data set using ggplot.
#'
#' @param data an object of class "data.frame" containing just numerical columns.
#' @param columns an object of class "numeric" containing the list of columns
#' that you want in your scatterplot.
#' @param dependentVariable an object of class "numeric", "factor" or "integer" is
#' a list of values containig the dependent variable.
#' @param dependentVariableName is an optional parameter. It's an string that
#' contains the name of your dependent variable.
#' @param pointSize is an optional parameter of class numeric with a single value
#' that represent the point size of plot.
#' @param alphaPoint is an optional parameter of class numeric with a single value
#' that represent the alpha of points in the plot.
#' @param colours is an optional parameter of class character with a list of colours
#' to use in the plot. The default value for continuos dependent variable is
#' c("darkred", "yellow", "darkgreen") and for categorical dependent variable are
#' the default colours defined by ggplot.
#' @seealso makePairs
#' @source https://gastonsanchez.wordpress.com/2012/08/27/scatterplot-matrices-with-ggplot/
#' @examples
#' #Example 1
#' iris.x <- iris[,1:4] # These are the independent variables
#' Species <- iris[,5] # This is the dependent variable
#'
#' # A Scatterplot of all columns
#' ScatterplotMatrix(iris.x, c(1,2,3,4), Species, "Species")
#' # A Scatterplot of somes columns and different point size and alpha point
#' ScatterplotMatrix(iris.x, c(2,4), Species, "Species", 2, 1)
#' # A Scatterplot with a different colours palette
#' myPalette <- c("darkolivegreen4", "goldenrod1", "dodgerblue4")
#' ScatterplotMatrix(iris.x, c(2,4), Species, "Species", 2, 1, colours = myPalette)
#'
#'
#' #Example 2
#' # Getting a clean data set (without missing values)
#' cars <- read.csv("https://dl.dropboxusercontent.com/u/12599702/autosclean.csv", sep = ";", dec = ",")
#' cars.x <- cars[,1:16] # These are the independent variables
#' cars.y <- cars[,17] # This is the dependent variable
#'
#' # A Scatterplot of some columns
#' ScatterplotMatrix(cars.x, seq(3, 8, 1), cars.y, "Price")
#' # A Scatterplot of somes columns and different point size and alpha point
#' ScatterplotMatrix(cars.x, c(2,4), cars.y, "Price", 2, 1)
#' # A Scatterplot with a different colours palette
#' myPalette <- c("darkolivegreen4", "goldenrod1", "dodgerblue4")
#' ScatterplotMatrix(cars.x, c(2,4), cars.y, "Price", 1.5, 1, colours = myPalette)
#'
#' #Example 3
#' # Getting a clean data set (without missing values)
#' cars <- read.csv("https://dl.dropboxusercontent.com/u/12599702/autosclean.csv", sep = ";", dec = ",")
#' cars.x <- cars[,1:16] # These are the independent variables
#' cars.y <- cars[,17] # This is the dependent variable
#'
#' # Performing prcomp
#' cars.pca <- prcomp(cars.x, center = TRUE, scale. = TRUE)
#'
#' # A Scatterplot of some columns of principal components
#' ScatterplotMatrix(as.data.frame(cars.pca$x), seq(1, 4, 1), cars.y, "Price")
ScatterplotMatrix <- function(data, columns, dependentVariable, dependentVariableName, pointSize, alphaPoint, colours){
if (missing(data)) {
stop("Need to specify data!")
}
if (class(data) != "data.frame") {
stop("data must be a data.frame class!")
}
if (missing(columns)) {
stop("Need to specify columns!")
}
if (!(class(columns) == "numeric" || class(columns) == "integer")) {
stop("columns must be a numeric or integer class!")
}
if (missing(dependentVariable)) {
stop("Need to specify dependentVariable!")
}
if (!(class(dependentVariable) == "numeric" || class(dependentVariable) == "factor" || class(dependentVariable) == "integer")) {
stop("dependentVariable must be a numeric, factor or integer class!")
}
if (missing(dependentVariableName)) {
dependentVariableName <- "Dependent Variable"
}
if (class(dependentVariableName) != "character") {
stop("dependentVariableName must be a character class!")
}
if (missing(pointSize)) {
pointSize <- 1
}
if (class(pointSize) != "numeric") {
stop("pointSize must be a numeric class!")
}
if (missing(alphaPoint)) {
alphaPoint <- 0.5
}
if (class(alphaPoint) != "numeric") {
stop("alphaPoint must be a numeric class!")
}
if (missing(colours)) {
if (class(dependentVariable) == "numeric" || class(dependentVariable) == "integer") {
colours <- c("darkred", "yellow", "darkgreen")
}
}
#All parameters are OK!
# expand data frame for pairs plot
subData <- as.data.frame(data[,columns])
gg1 <- makePairs(subData)
#New data frame mega Data from..to
mega_Data <- data.frame(gg1$all, DependentVariable = rep(dependentVariable, length = nrow(gg1$all)))
DependentVariable <- rep(dependentVariable, length = nrow(gg1$all))
# pairs plot
if (class(dependentVariable) == "numeric" || class(dependentVariable) == "integer") {
min <- min(dependentVariable)
max <- max(dependentVariable)
p <- ggplot(mega_Data, aes_string(x = "x", y = "y")) +
facet_grid(xvar ~ yvar, scales = "free") +
geom_point(aes(colour = DependentVariable), na.rm = TRUE, alpha = alphaPoint, size = pointSize) +
stat_density(aes(x = x, y = ..scaled.. * diff(range(x)) + min(x)),
data = gg1$densities, position = "identity",
colour = "dodgerblue4", geom = "line", size = 1, alpha = 0.5) +
scale_color_gradientn(name = dependentVariableName,
colours = colours, breaks = c(min, max),
labels = c(min, max)) + #set the pallete
theme(panel.grid.minor = element_blank(), #remove gridlines
legend.position = "bottom", #legend at the bottom
axis.title.x = element_blank(), #remove x label
axis.title.y = element_blank() #remove y label
)#end theme
}
else {
if (missing(colours)) {
p <- ggplot(mega_Data, aes_string(x = "x", y = "y")) +
facet_grid(xvar ~ yvar, scales = "free") +
geom_point(aes(colour = DependentVariable), na.rm = TRUE, alpha = alphaPoint, size = pointSize) +
stat_density(aes(x = x, y = ..scaled.. * diff(range(x)) + min(x)),
data = gg1$densities, position = "identity",
colour = "dodgerblue4", geom = "line", size = 1, alpha = 0.5) +
scale_color_discrete(name = dependentVariableName) +
guides(colour = guide_legend(override.aes = list(alpha = 1))) +
theme(panel.grid.minor = element_blank(), #remove gridlines
legend.position = "bottom", #legend at the bottom
axis.title.x = element_blank(), #remove x label
axis.title.y = element_blank() #remove y label
)#end theme
} else {
p <- ggplot(mega_Data, aes_string(x = "x", y = "y")) +
facet_grid(xvar ~ yvar, scales = "free") +
geom_point(aes(colour = DependentVariable), na.rm = TRUE, alpha = alphaPoint, size = pointSize) +
stat_density(aes(x = x, y = ..scaled.. * diff(range(x)) + min(x)),
data = gg1$densities, position = "identity",
colour = "dodgerblue4", geom = "line", size = 1, alpha = 0.5) +
scale_color_manual(values = colours) +
guides(colour = guide_legend(override.aes = list(alpha = 1))) +
theme(panel.grid.minor = element_blank(), #remove gridlines
legend.position = "bottom", #legend at the bottom
axis.title.x = element_blank(), #remove x label
axis.title.y = element_blank() #remove y label
)#end theme
}
}
return (p)
}
#' Parallel Plot (Plot)
#'
#' Generate a plot of the columns of a data set for all or a range of instances. In
#' some cases this is useful to identify some patron.
#'
#' @param data an object of class "data.frame" containing just numerical columns.
#' @param rows an object of class "numeric" containing the list of rows
#' that you want in your parallel plot.
#' @param columns an object of class "numeric" containing the list of columns
#' that you want in your parallel plot.
#' @param dependentVariable an object of class "numeric", "factor" or "integer" is
#' a list of values containig the dependent variable.
#' @param dependentVariableName is an optional parameter. It's an string that
#' contains the name of your dependent variable.
#' @param lineSize is an optional parameter of class numeric with a single value
#' that represent the line size of plot.
#' @param alphaLine is an optional parameter of class numeric with a single value
#' that represent the alpha of lines in the plot.
#' @param x_lab a boolean that represent if you want or not the x axis scale. In
#' some cases, when you have many columns the plot could be ugly! The default value
#' is False.
#' @param colours is an optional parameter of class character with a list of colours
#' to use in the plot. The default value for continuos dependent variable is
#' c("darkred", "yellow", "darkgreen") and for categorical dependent variable are
#' the default colours defined by ggplot.
#'
#' @examples
#' #Example 1
#' iris.x <- iris[,1:4] # These are the independent variables
#' Species <- iris[,5] # This is the dependent variable
#'
#' # A ParallelPlot of all rows and all columns
#' ParallelPlot(iris.x, seq(1,nrow(iris.x),1), seq(1,ncol(iris.x),1), Species, "Species", 1, 0.5, TRUE)
#' # A ParallelPlot of all rows and some columns
#' ParallelPlot(iris.x, seq(1,nrow(iris.x),1), c(3,4), Species, "Species", 1, 0.5, TRUE)
#' # A ParallelPlot with a different colours palette
#' myPalette <- c("darkolivegreen4", "goldenrod1", "dodgerblue4")
#' ParallelPlot(iris.x, seq(1,nrow(iris.x),1), seq(1,ncol(iris.x),1), Species, "Species", 1, 0.5, TRUE, colours = myPalette)
#'
#'
#' #Example 2
#' # Getting a clean data set (without missing values)
#' cars <- read.csv("https://dl.dropboxusercontent.com/u/12599702/autosclean.csv", sep = ";", dec = ",")
#' cars.x <- cars[,1:16] # These are the independent variables
#' cars.y <- cars[,17] # This is the dependent variable
#'
#' # A ParallelPlot of all rows and all columns
#' ParallelPlot(cars.x, seq(1,nrow(cars.x),1), seq(1,ncol(cars.x),1), cars.y, "Price", 1, 0.5, TRUE)
#' # A ParallelPlot of all rows and some columns
#' ParallelPlot(cars.x, seq(1,nrow(cars.x),1), c(1,2,5,8,13,14), cars.y, "Price", 1, 0.8, TRUE)
#' # A ParallelPlot with a different colours palette
#' myPalette <- c("darkolivegreen4", "goldenrod1", "dodgerblue4")
#' ParallelPlot(cars.x, seq(1,nrow(cars.x),1), c(1,2,5,8,13,14), cars.y, "Price", 1, 0.8, TRUE, colours = myPalette)
ParallelPlot <- function(data, rows, columns, dependentVariable, dependentVariableName, lineSize, alphaLine, x_lab, colours) {
if (missing(data)) {
stop("Need to specify data!")
}
if (class(data) != "data.frame") {
stop("data must be a data.frame class!")
}
if (missing(rows)) {
stop("Need to specify rows!")
}
if (!(class(rows) == "numeric" || class(rows) == "integer")) {
stop("rows must be a numeric or integer class!")
}
if (missing(columns)) {
stop("Need to specify columns!")
}
if (!(class(columns) == "numeric" || class(columns) == "integer")) {
stop("columns must be a numeric or integer class!")
}
if (missing(dependentVariable)) {
stop("Need to specify dependentVariable!")
}
if (!(class(dependentVariable) == "numeric" || class(dependentVariable) == "factor" || class(dependentVariable) == "integer")) {
stop("dependentVariable must be a numeric, factor or integer class!")
}
if (missing(dependentVariableName)) {
dependentVariableName <- "Dependent Variable"
}
if (class(dependentVariableName) != "character") {
stop("dependentVariableName must be a character class!")
}
if (missing(lineSize)) {
lineSize <- 1
}
if (class(lineSize) != "numeric") {
stop("lineSize must be a numeric class!")
}
if (missing(alphaLine)) {
alphaLine <- 0.9
}
if (class(alphaLine) != "numeric") {
stop("alphaLine must be a numeric class!")
}
if (missing(x_lab)) {
x_lab = FALSE
}
if(class(x_lab) != "logical") {
stop("x_lab must be a logical class!")
}
if (missing(colours)) {
if (class(dependentVariable) == "numeric" || class(dependentVariable) == "integer") {
colours <- c("darkred", "yellow", "darkgreen")
}
}
#All parameters are OK!
x_name = "Wavelength"
subData <- data[rows,columns]
dependentVariable <- dependentVariable[rows]
rowsNum <- nrow(subData)
x <- seq(from = 1, to = rowsNum, length.out = rowsNum)
data <- data.frame(x, subData)
dataPlot <- melt(data, id = "x")
dataPlot <- data.frame(dataPlot, dependentVariable)
if (class(dependentVariable) == "numeric" || class(dependentVariable) == "integer") {
if (x_lab) {
min <- min(dependentVariable)
max <- max(dependentVariable)
p <- ggplot(dataPlot, aes(variable, value, group = x, colour = dependentVariable)) +
geom_line(size = lineSize, alpha = alphaLine) +
scale_color_gradientn(name = dependentVariableName,
colours = colours, breaks = c(min, max),
labels = c(min, max)) + #set the pallete
xlab(x_name) + ylab("Values") +
theme(legend.position = "bottom" #legend at the bottom
)#end theme
}
else {
min <- min(dependentVariable)
max <- max(dependentVariable)
p <- ggplot(dataPlot, aes(variable, value, group = x, colour = dependentVariable)) +
geom_line(size = lineSize, alpha = alphaLine) +
scale_color_gradientn(name = dependentVariableName,
colours = colours, breaks = c(min, max),
labels = c(min, max)) + #set the pallete
scale_x_discrete(breaks = c()) +
xlab(x_name) + ylab("Values") +
theme(legend.position = "bottom" #legend at the bottom
)#end theme
}
}
else {
if (x_lab) {
if (missing(colours)) {
p <- ggplot(dataPlot, aes(variable, value, group = x, colour = dependentVariable)) +
geom_line(size = lineSize, alpha = alphaLine) +
scale_color_discrete(name = dependentVariableName) +
xlab(x_name) + ylab("Values") +
guides(colour = guide_legend(override.aes = list(alpha = 1))) +
theme(legend.position = "bottom" #legend at the bottom
)#end theme +
} else {
p <- ggplot(dataPlot, aes(variable, value, group = x, colour = dependentVariable)) +
geom_line(size = lineSize, alpha = alphaLine) +
scale_color_manual(values = colours) +
xlab(x_name) + ylab("Values") +
guides(colour = guide_legend(override.aes = list(alpha = 1))) +
theme(legend.position = "bottom" #legend at the bottom
)#end theme
}
}
else {
if (missing(colours)) {
p <- ggplot(dataPlot, aes(variable, value, group = x, colour = dependentVariable)) +
geom_line(size = lineSize, alpha = alphaLine) +
scale_color_discrete(name = dependentVariableName) +
scale_x_discrete(breaks = c()) +
xlab(x_name) + ylab("Values") +
guides(colour = guide_legend(override.aes = list(alpha = 1))) +
theme(legend.position = "bottom" #legend at the bottom
)#end theme
} else {
p <- ggplot(dataPlot, aes(variable, value, group = x, colour = dependentVariable)) +
geom_line(size = lineSize, alpha = alphaLine) +
scale_color_manual(values = colours) +
scale_x_discrete(breaks = c()) +
xlab(x_name) + ylab("Values") +
guides(colour = guide_legend(override.aes = list(alpha = 1))) +
theme(legend.position = "bottom" #legend at the bottom
)#end theme
}
}
}
return (p)
}
#' Density Plot (Plot)
#'
#' Generate a density plot for a specific column of the data.
#'
#' @param data an object of class data frame with the data.
#' @param col an integer that specify the column that you want for make the plot.
#'
#' @seealso http://www.rdatamining.com/examples/outlier-detection
#'
#' @examples
#' #Example 1
#' #install.packages("Rlof")
#' library(Rlof) #for outlier detection
#'
#' iris.x <- iris[,1:4] # These are the independent variables
#' Species <- iris[,5] # This is the dependent variable
#'
#' DensityPlot(iris.x,1)
#'
#'
#' #Example 2
#' #install.packages("Rlof")
#' library(Rlof) #Outlier detection library
#'
#' iris.x <- iris[,1:4] # These are the independent variables
#' Species <- iris[,5] # This is the dependent variable
#'
#' outlier.scores <- lof(iris.x, k = 5) #applying outlier detection
#' outlier.scores <- data.frame(outlier.scores)
#' DensityPlot(outlier.scores, 1) #Generating a plot of outliers scores
#'
#'
#' #Example 3
#' #install.packages("Rlof")
#' library(Rlof) #Outlier detection library
#'
#' iris.x <- iris[,1:4] # These are the independent variables
#' Species <- iris[,5] # This is the dependent variable
#'
#' outlier.scores <- lof(iris.x, k = c(5:10)) #applying outlier detection
#' mean <- rowMeans(outlier.scores) #Calculating the mean of every execution
#' outlier.scores <- data.frame(outlier.scores, mean) #adding mean to data frame
#' DensityPlot(outlier.scores, ncol(outlier.scores)) #Generating a plot of outliers scores
#'
#'
#' #Example 4
#' #install.packages("Rlof")
#' library(Rlof) #Outlier detection library
#' library(plyr)
#' # Getting a data set without missing values
#' cars <- read.csv("https://dl.dropboxusercontent.com/u/12599702/autosclean.csv", sep = ";", dec = ",")
#'
#' cars.x <- cars[,1:16] # These are the independent variables
#' cars.y <- cars[,17] # This is the dependent variable
#'
#' outlier.scores <- lof(cars.x, k = c(5:10)) #applying outlier detection
#' mean <- rowMeans(outlier.scores) #Calculating the mean of every execution
#' outlier.scores <- data.frame(outlier.scores, mean) #adding mean to data frame
#' DensityPlot(outlier.scores, ncol(outlier.scores)) #Generating a plot of outliers scores
#'
#' aux <- outlier.scores[,7]>1.7 #1.7 is the threshold selected
#' count(aux)[2,2] #Number of outliers found
#' outliers <- order(outlier.scores[,7], decreasing=T)[1:count(aux)[2,2]] #Getting the values that are on the threshold
#' Score <- outlier.scores[outliers,7] #Getting outliers scores
#' outliers <- data.frame(outliers,Score)
#' names(outliers) <- c("Position","Score")
#' View(outliers)
#'
#' auxOutliers <- outlier.scores[-outliers[1:3,1],] #Eliminating the 3 most remote instances!
#' DensityPlot(auxOutliers, ncol(outlier.scores)) #Generating a plot of outliers scores
DensityPlot <- function(data, col) {
if (missing(data)) {
stop("Need to specify data!")
}
if (class(data) != "data.frame") {
stop("data must be a data frame class!")
}
if (missing(col)) {
stop("Need to specify col!")
}
if (!(class(col) == "numeric" || class(col) == "integer")) {
stop("col must be a numeric or integer class!")
}
if (col > ncol(data)) {
stop("Col value must be less than ncol of data!")
}
#All parameters are OK!
names(data)[col] <- "mean"
p <- ggplot(data, aes(x = mean)) +
geom_density(colour="darkgreen", fill="darkgreen", alpha=0.3) +
xlab("Values") +
ylab("Density")
return (p)
}
#' Plot PC 3D (3DPlot)
#'
#' Generate a 3D plot Generates a 3D graphic for a set of 3 columns of the data set.
#'
#' @param data an object of class data frame with the data.
#' @param columns an object of class "numeric" containing the list of columns
#' that you want in your parallel plot.
#' @param dependentVariable an object of class "numeric", "factor" or "integer" is
#' a list of values containig the dependent variable.
#'
#' @examples
#' #Example 1
#' iris.x <- iris[,1:4] # These are the independent variables
#' Species <- iris[,5] # This is the dependent variable
#'
#' # 3D Plot of 3 first columns of data set
#' Plot3D(iris.x, c(1,2,3), Species)
#'
#'
#' #Example 2
#' iris.x <- iris[,1:4] # These are the independent variables
#' Species <- iris[,5] # This is the dependent variable
#'
#' ir.pca <- prcomp(iris.x, center = TRUE, scale. = TRUE) # performin prcomp
#'
#' # 3D Plot of 3 first columns of data set
#' Plot3D(as.data.frame(ir.pca$x), c(1,2,3), Species)
#'
#'
#' #Example 3
#' # Getting a clean data set (without missing values)
#' cars <- read.csv("https://dl.dropboxusercontent.com/u/12599702/autosclean.csv", sep = ";", dec = ",")
#' cars.x <- cars[,1:16] # These are the independent variables
#' cars.y <- cars[,17] # This is the dependent variable
#'
#' # 3D Plot of 3 first columns of data set
#' Plot3D(cars.x, c(1,2,3), cars.y)
#'
#'
#' #Example 4
#' # Getting a clean data set (without missing values)
#' cars <- read.csv("https://dl.dropboxusercontent.com/u/12599702/autosclean.csv", sep = ";", dec = ",")
#' cars.x <- cars[,1:16] # These are the independent variables
#' cars.y <- cars[,17] # This is the dependent variable
#'
#' cars.pca <- prcomp(cars.x, center = TRUE, scale. = TRUE) # performin prcomp
#'
#' # 3D Plot of 3 first columns of data set
#' Plot3D(as.data.frame(cars.pca$x), c(1,2,3), cars.y)
Plot3D <- function(data, columns, dependentVariable){
if (missing(data)) {
stop("Need to specify data!")
}
if (class(data) != "data.frame") {
stop("data must be a data.frame class!")
}
if (missing(columns)) {
stop("Need to specify columns!")
}
if (!(class(columns) == "numeric" || class(columns) == "integer")) {
stop("columns must be a numeric or integer class!")
}
if (length(columns) != 3) {
stop("The number of selected columns must be 3!")
}
if (missing(dependentVariable)) {
stop("Need to specify dependentVariable!")
}
if (!(class(dependentVariable) == "numeric" || class(dependentVariable) == "factor" || class(dependentVariable) == "integer")) {
stop("dependentVariable must be a numeric, factor or integer class!")
}
#All parameters are OK!
subData<-data[,columns]
col1 <- subData[,1]
col2 <- subData[,2]
col3 <- subData[,3]
x_lab <- colnames(subData)[1]
y_lab <- colnames(subData)[2]
z_lab <- colnames(subData)[3]
if (class(dependentVariable) == "numeric" || class(dependentVariable) == "integer") {
cols <- myColorRamp(c("darkred", "yellow", "darkgreen"), dependentVariable)
plot3d(x = col1, y = col2, z = col3, col = cols, size = "4", xlab = x_lab, ylab = y_lab, zlab = z_lab)
}
else {
cols <- myColorRamp(c("darkred", "yellow", "darkgreen"), as.numeric(dependentVariable))
plot3d(x = col1, y = col2, z = col3, col = cols, size = "4", xlab = x_lab, ylab = y_lab, zlab = z_lab)
}
}
#' Simple Plot of 2 columns (Plot)
#'
#' Generate a plot of 2 columns of data set using ggplot. You must indicate which
#' columns you want in the graph.
#'
#' @param data an object of class data frame with the data.
#' @param DependentVariable an object of class "numeric", "factor" or "integer" is
#' a list of values containig the dependent variable.
#' @param x_axis an integer that represent the number of the column that you want
#' in your x axis.
#' @param y_axis an integer that represent the number of the column that you want
#' in your y axis.
#' @param dependentVariableName is an optional parameter. It's an string that
#' contains the name of your dependent variable.
#' @param pointSize is an optional parameter of class numeric with a single value
#' that represent the point size of plot.
#' @param alphaPoint is an optional parameter of class numeric with a single value
#' that represent the alpha of points in the plot.
#' @param colours is an optional parameter of class character with a list of colours
#' to use in the plot. The default value for continuos dependent variable is
#' c("darkred", "yellow", "darkgreen") and for categorical dependent variable are
#' the default colours defined by ggplot.
#'
#' @seealso elbowPlot
#'
#' @examples
#' #Example 1
#' iris.x <- iris[,1:4] # These are the independent variables
#' Species <- iris[,5] # This is the dependent variable
#'
#' # Plot of first 2 columns of data set
#' simplePlot(iris.x, Species, 1, 2, "Species", 2, 0.9)
#' # A plot with a different colours palette
#' myPalette <- c("darkolivegreen4", "goldenrod1", "dodgerblue4")
#' simplePlot(iris.x, Species, 1, 2, "Species", 2, 0.9, colours = myPalette)
#'
#'
#' #Example 2
#' iris.x <- iris[,1:4] # These are the independent variables
#' Species <- iris[,5] # This is the dependent variable
#'
#' ir.pca <- prcomp(iris.x, center = TRUE, scale. = TRUE) #performing prcomp
#'
#' # Plot of first 2 columns of principal components
#' simplePlot(as.data.frame(ir.pca$x), Species, 1, 2, "Species", 2, 0.9)
#' # A plot with a different colours palette
#' myPalette <- c("darkolivegreen4", "goldenrod1", "dodgerblue4")
#' simplePlot(as.data.frame(ir.pca$x), Species, 1, 2, "Species", 2, 0.9, colours = myPalette)
#'
#'
#' #Example 3
#' # Getting a clean data set (without missing values)
#' cars <- read.csv("https://dl.dropboxusercontent.com/u/12599702/autosclean.csv", sep = ";", dec = ",")
#' cars.x <- cars[,1:16] # These are the independent variables
#' cars.y <- cars[,17] # This is the dependent variable
#'
#' cars.pca <- prcomp(cars.x, center = TRUE, scale. = TRUE) #performing prcomp
#'
#' # Plot of first 2 columns of principal components
#' simplePlot(as.data.frame(cars.pca$x), cars.y, 1, 2, "Price", 2, 0.9)
#' # A plot with a different colours palette
#' myPalette <- c("darkolivegreen4", "goldenrod1", "dodgerblue4")
#' simplePlot(as.data.frame(cars.pca$x), cars.y, 1, 2, "Price", 2, 0.9, colours = myPalette)
simplePlot <- function(data, DependentVariable, x_axis, y_axis, dependentVariableName, pointSize, alphaPoint, colours) {
if (missing(data)) {
stop("Need to specify data!")
}
if (class(data) != "data.frame") {
stop("data must be a data.frame class!")
}
if (missing(DependentVariable)) {
stop("Need to specify DependentVariable!")
}
if (!(class(DependentVariable) == "numeric" || class(DependentVariable) == "factor" || class(DependentVariable) == "integer")) {
stop("DependentVariable must be a numeric, factor or integer class!")
}
if (missing("x_axis")) {
stop("Need to specify x_axis!")
}
if (!(class(x_axis) == "numeric" || class(x_axis) == "integer")) {
stop("x_axis must be a numeric or integer class!")
}
if (missing("y_axis")) {
stop("Need to specify y_axis!")
}
if (!(class(y_axis) == "numeric" || class(y_axis) == "integer")) {
stop("y_axis must be a numeric or integer class!")
}
if (missing(dependentVariableName)) {
dependentVariableName <- "Dependent Variable"
}
if (class(dependentVariableName) != "character") {
stop("dependentVariableName must be a character class!")
}
if (missing(pointSize)) {
pointSize <- 1
}
if (class(pointSize) != "numeric") {
stop("pointSize must be a numeric class!")
}
if (missing(alphaPoint)) {
alphaPoint <- 0.5
}
if (class(alphaPoint) != "numeric") {
stop("alphaPoint must be a numeric class!")
}
if (missing(colours)) {
if (class(DependentVariable) == "numeric" || class(DependentVariable) == "integer") {
colours <- c("darkred", "yellow", "darkgreen")
}
}
#All parameters are OK!
subData <- data.frame(data[,x_axis], data[,y_axis], DependentVariable)
x_axis <- colnames(data)[x_axis]
y_axis <- colnames(data)[y_axis]
names(subData) <- c(x_axis, y_axis, "DependentVariable")
if (class(DependentVariable) == "numeric" || class(DependentVariable) == "integer") {
min <- min(DependentVariable)
max <- max(DependentVariable)
p <- ggplot(subData, aes_string(x = x_axis, y = y_axis)) +
geom_point(aes(colour = DependentVariable), na.rm = TRUE, alpha = alphaPoint, size = pointSize) +
scale_color_gradientn(name = dependentVariableName,
colours = colours, breaks = c(min, max),
labels = c(min, max)) + #set the pallete
theme(panel.grid.minor = element_blank(), #remove gridlines
legend.position = "bottom" #legend at the bottom
) + #end theme
xlab(x_axis) + ylab(y_axis)
}
else {
if (missing(colours)) {
p <- ggplot(subData, aes_string(x = x_axis, y = y_axis)) +
geom_point(aes(colour = DependentVariable), na.rm = TRUE, alpha = alphaPoint, size = pointSize) +
scale_color_discrete(name = dependentVariableName) +
guides(colour = guide_legend(override.aes = list(alpha = 1))) +
theme(panel.grid.minor = element_blank(), #remove gridlines
legend.position = "bottom" #legend at the bottom
) + #end theme
xlab(x_axis) + ylab(y_axis)
} else {
p <- ggplot(subData, aes_string(x = x_axis, y = y_axis)) +
geom_point(aes(colour = DependentVariable), na.rm = TRUE, alpha = alphaPoint, size = pointSize) +
scale_color_manual(values = colours) +
guides(colour = guide_legend(override.aes = list(alpha = 1))) +
theme(panel.grid.minor = element_blank(), #remove gridlines
legend.position = "bottom" #legend at the bottom
) + #end theme
xlab(x_axis) + ylab(y_axis)
}
}
return (p)
}
#' Ridge Plot
#'
#' Plot the cross-validation curve produced by cv.glmnet using ggplot
#'
#' @param ridgeModelCV an object of class cv.glmnet that contain the cross-validation
#' information for generate a curve.
#' @param pointSize is an optional parameter of class numeric with a single value
#' that represent the point size of plot. The default value is 1.5
#' @param alphaPoint is an optional parameter of class numeric with a single value
#' that represent the alpha of points in the plot. The default value is 0.8
#' @param colourPoint is an optional parameter of class character with a single
#' colour. The default value is "darkred".
#' @param colourLine is an optional parameter of class character with a single
#' colour. The default value is "dodgerblue4".
#' @param errorMode is an optional parameter of class character that defined if the
#' error will be displayed using "errorbar" or "ribbon". The default value is "errorbar"
#' @param colourError is an optional parameter of class character with a single
#' colour. The default value is "#A9A9A9".
#'
#' @examples
#' #Example 1
#' #install.packages("MASS")
#' library(MASS) #for ridge regression
#' #install.packages("glmnet")
#' library(glmnet) #for parameter optimization
#' iris.x <- iris[,1:3] # These are the independent variables
#' iris.y <- iris[,4] # This is the dependent variable
#'
#' #We perform the ridge method
#' grid <- 10^seq(10, -2, length = 100) #Define a large grid for lambda values
#' set.seed(2015)
#' ridge <- cv.glmnet(as.matrix(iris.x), iris.y, alpha = 0, lambda = grid) #alpha = 0 for ridge, alpha = 1 for lasso
#' RidgePlot(ridge)
#'
#' bestLambda <- ridge$lambda.min # The optimal lambda
#' ridge.final <- lm.ridge(y ~ ., data = X, lambda = bestLambda)
#'
#'
#' #Example 2
#' #install.packages("MASS")
#' library(MASS) #for ridge regression
#' #install.packages("glmnet")
#' library(glmnet) #for parameter optimization
#' # Getting a clean data set (without missing values)
#' cars <- read.csv("https://dl.dropboxusercontent.com/u/12599702/autosclean.csv", sep = ";", dec = ",")
#' cars.x <- cars[,1:16] # These are the independent variables
#' cars.y <- cars[,17] # This is the dependent variable
#'
#' #We perform the ridge method
#' grid <- 10^seq(10, -2, length = 100) #Define a large grid for lambda values
#' set.seed(2015)
#' ridge <- cv.glmnet(as.matrix(cars.x), cars.y, alpha = 0, lambda = grid) #alpha = 0 for ridge, alpha = 1 for lasso
#' RidgePlot(ridge)
#' RidgePlot(ridge, errorMode = "ribbon") #We change the plot with geom_ribbon for errors
#'
#' bestLambda <- ridge$lambda.min # The optimal lambda
#' ridge.final <- lm.ridge(y ~ ., data = X, lambda = bestLambda)
RidgePlot <- function(ridgeModelCV, pointSize, alphaPoint, colourPoint, colourLine, errorMode, colourError) {
if (missing(ridgeModelCV)) {
stop("Need to specify ridgeModelCV!")
}
if (class(ridgeModelCV) != "cv.glmnet") {
stop("ridgeModelCV must be a cv.glmnet class!")
}
if (missing(pointSize)) {
pointSize <- 1.5
}
if (class(pointSize) != "numeric") {
stop("pointSize must be a numeric class!")
}
if (missing(alphaPoint)) {
alphaPoint <- 0.8
}
if (class(alphaPoint) != "numeric") {
stop("alphaPoint must be a numeric class!")
}
if (missing(colourPoint)) {
colourPoint <- "darkred"
}
if (class(colourPoint) != "character") {
stop("colourPoint must be a character class!")
}
if (missing(colourLine)) {
colourLine <- "dodgerblue4"
}
if (class(colourLine) != "character") {
stop("colourLine must be a character class!")
}
if (missing(errorMode)) {
errorMode <- "errorbar"
}
if (class(errorMode) != "character") {
stop("errorMode must be a character class!")
} else {
if (!(errorMode== "errorbar" || errorMode == "ribbon")) {
stop("errorMode must be a errorbar or ribbon!")
}
}
if (missing(colourError)) {
colourError <- "#A9A9A9"
}
if (class(colourError) != "character") {
stop("colourError must be a character class!")
}
tidied_cv <- tidy(ridgeModelCV)
glance_cv <- glance(ridgeModelCV)
tidied_cv <- data.frame(tidied_cv, log(tidied_cv$lambda))
names(tidied_cv) <- c(names(tidied_cv)[1:6], "logLambda")
if (errorMode == "errorbar") {
p <- ggplot(tidied_cv, aes(logLambda, estimate)) +
xlab("log(Lambda)") + ylab("MSE") +
geom_errorbar(aes(ymin = conf.low, ymax = conf.high), colour = colourError, width = .5) +
geom_vline(xintercept = log(glance_cv$lambda.min), lty = 2, colour = colourLine) +
geom_vline(xintercept = log(glance_cv$lambda.1se), lty = 2, colour = colourLine) +
geom_point(colour = colourPoint, size = pointSize, alpha = alphaPoint)
} else {
p <- ggplot(tidied_cv, aes(logLambda, estimate)) +
xlab("log(Lambda)") + ylab("MSE") +
geom_ribbon(aes(ymin = conf.low, ymax = conf.high), alpha = .25, colour = colourError) +
geom_vline(xintercept = log(glance_cv$lambda.min), lty = 2, colour = colourLine) +
geom_vline(xintercept = log(glance_cv$lambda.1se), lty = 2, colour = colourLine) +
geom_point(colour = colourPoint, size = pointSize, alpha = alphaPoint)
}
return (p)
}
mariytu/RegressionLibs documentation built on May 21, 2019, 11:47 a.m.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Defines functions elbowPlot ScatterplotMatrix ParallelPlot DensityPlot Plot3D simplePlot RidgePlot
Documented in DensityPlot elbowPlot ParallelPlot Plot3D ScatterplotMatrix simplePlot
#' Elbow Plot for PCA (Plot)
#'
#' Generate a plot of 10 first variances of Principal Components. This is useful to
#' determinate which are the most important components.
#'
#' @param data.pca a list with class "prcomp" containing all principal components
#' calculated.
#' @seealso CalculateVariance, plotPC
#' @examples
#' #Example 1
#' iris.x <- iris[,1:4] # These are the independent variables
#' # We know that there are no missing values in the data set
#'
#' # performing prcomp
#' ir.pca <- prcomp(iris.x, center = TRUE, scale. = TRUE)
#'
#' # Generating elbow plot to detect the most important principal components
#' elbowPlot(ir.pca)
#'
#'
#' #Example 2
#' # Getting a clean data set (without missing values)
#' cars <- read.csv("https://dl.dropboxusercontent.com/u/12599702/autosclean.csv", sep = ";", dec = ",")
#' cars.x <- cars[,1:16] # These are the independent variables
#'
#' # Performing prcomp
#' cars.pca <- prcomp(cars.x, center = TRUE, scale. = TRUE)
#'
#' # Generating elbow plot to detect the most important principal components
#' elbowPlot(cars.pca)
elbowPlot <- function(data.pca) {
if (missing(data.pca)) {
stop("Need to specify data.pca!")
}
if (class(data.pca) != "prcomp") {
stop("data.pca must be a prcomp class!")
}
#All parameters are OK!
rowsData <- length(data.pca$sdev)
seqRow <- seq(from = 1, to = rowsData, length.out = rowsData)
dataPlot <- data.frame(seqRow, data.pca$sdev)
names(dataPlot) <- c("PCA", "Variances")
if (nrow(dataPlot)>10) {
dataPlot <- dataPlot[1:10,]
}
dataPlot <- CalculateVariance(dataPlot, 2)
p <- ggplot(data = dataPlot, aes(x = PCA, y = Variances, group = 1)) +
geom_line(colour = "dodgerblue4", alpha = 0.5, size = 1) +
geom_point(colour = "dodgerblue4", size = 2, alpha = 0.5) +
expand_limits(y = 0) +
xlab("PCs") + ylab("Variances") +
scale_x_continuous(breaks = dataPlot$PCA) +
theme(panel.grid.minor = element_blank(), #remove gridlines
legend.position = "bottom" #legend at the bottom
)#end theme
return (p)
}
#' Scatterplot Matrix (Plot)
#'
#' Generate a Scatterplot Matrix of some columns of data set using ggplot.
#'
#' @param data an object of class "data.frame" containing just numerical columns.
#' @param columns an object of class "numeric" containing the list of columns
#' that you want in your scatterplot.
#' @param dependentVariable an object of class "numeric", "factor" or "integer" is
#' a list of values containig the dependent variable.
#' @param dependentVariableName is an optional parameter. It's an string that
#' contains the name of your dependent variable.
#' @param pointSize is an optional parameter of class numeric with a single value
#' that represent the point size of plot.
#' @param alphaPoint is an optional parameter of class numeric with a single value
#' that represent the alpha of points in the plot.
#' @param colours is an optional parameter of class character with a list of colours
#' to use in the plot. The default value for continuos dependent variable is
#' c("darkred", "yellow", "darkgreen") and for categorical dependent variable are
#' the default colours defined by ggplot.
#' @seealso makePairs
#' @source https://gastonsanchez.wordpress.com/2012/08/27/scatterplot-matrices-with-ggplot/
#' @examples
#' #Example 1
#' iris.x <- iris[,1:4] # These are the independent variables
#' Species <- iris[,5] # This is the dependent variable
#'
#' # A Scatterplot of all columns
#' ScatterplotMatrix(iris.x, c(1,2,3,4), Species, "Species")
#' # A Scatterplot of somes columns and different point size and alpha point
#' ScatterplotMatrix(iris.x, c(2,4), Species, "Species", 2, 1)
#' # A Scatterplot with a different colours palette
#' myPalette <- c("darkolivegreen4", "goldenrod1", "dodgerblue4")
#' ScatterplotMatrix(iris.x, c(2,4), Species, "Species", 2, 1, colours = myPalette)
#'
#'
#' #Example 2
#' # Getting a clean data set (without missing values)
#' cars <- read.csv("https://dl.dropboxusercontent.com/u/12599702/autosclean.csv", sep = ";", dec = ",")
#' cars.x <- cars[,1:16] # These are the independent variables
#' cars.y <- cars[,17] # This is the dependent variable
#'
#' # A Scatterplot of some columns
#' ScatterplotMatrix(cars.x, seq(3, 8, 1), cars.y, "Price")
#' # A Scatterplot of somes columns and different point size and alpha point
#' ScatterplotMatrix(cars.x, c(2,4), cars.y, "Price", 2, 1)
#' # A Scatterplot with a different colours palette
#' myPalette <- c("darkolivegreen4", "goldenrod1", "dodgerblue4")
#' ScatterplotMatrix(cars.x, c(2,4), cars.y, "Price", 1.5, 1, colours = myPalette)
#'
#' #Example 3
#' # Getting a clean data set (without missing values)
#' cars <- read.csv("https://dl.dropboxusercontent.com/u/12599702/autosclean.csv", sep = ";", dec = ",")
#' cars.x <- cars[,1:16] # These are the independent variables
#' cars.y <- cars[,17] # This is the dependent variable
#'
#' # Performing prcomp
#' cars.pca <- prcomp(cars.x, center = TRUE, scale. = TRUE)
#'
#' # A Scatterplot of some columns of principal components
#' ScatterplotMatrix(as.data.frame(cars.pca$x), seq(1, 4, 1), cars.y, "Price")
ScatterplotMatrix <- function(data, columns, dependentVariable, dependentVariableName, pointSize, alphaPoint, colours){
if (missing(data)) {
stop("Need to specify data!")
}
if (class(data) != "data.frame") {
stop("data must be a data.frame class!")
}
if (missing(columns)) {
stop("Need to specify columns!")
}
if (!(class(columns) == "numeric" || class(columns) == "integer")) {
stop("columns must be a numeric or integer class!")
}
if (missing(dependentVariable)) {
stop("Need to specify dependentVariable!")
}
if (!(class(dependentVariable) == "numeric" || class(dependentVariable) == "factor" || class(dependentVariable) == "integer")) {
stop("dependentVariable must be a numeric, factor or integer class!")
}
if (missing(dependentVariableName)) {
dependentVariableName <- "Dependent Variable"
}
if (class(dependentVariableName) != "character") {
stop("dependentVariableName must be a character class!")
}
if (missing(pointSize)) {
pointSize <- 1
}
if (class(pointSize) != "numeric") {
stop("pointSize must be a numeric class!")
}
if (missing(alphaPoint)) {
alphaPoint <- 0.5
}
if (class(alphaPoint) != "numeric") {
stop("alphaPoint must be a numeric class!")
}
if (missing(colours)) {
if (class(dependentVariable) == "numeric" || class(dependentVariable) == "integer") {
colours <- c("darkred", "yellow", "darkgreen")
}
}
#All parameters are OK!
# expand data frame for pairs plot
subData <- as.data.frame(data[,columns])
gg1 <- makePairs(subData)
#New data frame mega Data from..to
mega_Data <- data.frame(gg1$all, DependentVariable = rep(dependentVariable, length = nrow(gg1$all)))
DependentVariable <- rep(dependentVariable, length = nrow(gg1$all))
# pairs plot
if (class(dependentVariable) == "numeric" || class(dependentVariable) == "integer") {
min <- min(dependentVariable)
max <- max(dependentVariable)
p <- ggplot(mega_Data, aes_string(x = "x", y = "y")) +
facet_grid(xvar ~ yvar, scales = "free") +
geom_point(aes(colour = DependentVariable), na.rm = TRUE, alpha = alphaPoint, size = pointSize) +
stat_density(aes(x = x, y = ..scaled.. * diff(range(x)) + min(x)),
data = gg1$densities, position = "identity",
colour = "dodgerblue4", geom = "line", size = 1, alpha = 0.5) +
scale_color_gradientn(name = dependentVariableName,
colours = colours, breaks = c(min, max),
labels = c(min, max)) + #set the pallete
theme(panel.grid.minor = element_blank(), #remove gridlines
legend.position = "bottom", #legend at the bottom
axis.title.x = element_blank(), #remove x label
axis.title.y = element_blank() #remove y label
)#end theme
}
else {
if (missing(colours)) {
p <- ggplot(mega_Data, aes_string(x = "x", y = "y")) +
facet_grid(xvar ~ yvar, scales = "free") +
geom_point(aes(colour = DependentVariable), na.rm = TRUE, alpha = alphaPoint, size = pointSize) +
stat_density(aes(x = x, y = ..scaled.. * diff(range(x)) + min(x)),
data = gg1$densities, position = "identity",
colour = "dodgerblue4", geom = "line", size = 1, alpha = 0.5) +
scale_color_discrete(name = dependentVariableName) +
guides(colour = guide_legend(override.aes = list(alpha = 1))) +
theme(panel.grid.minor = element_blank(), #remove gridlines
legend.position = "bottom", #legend at the bottom
axis.title.x = element_blank(), #remove x label
axis.title.y = element_blank() #remove y label
)#end theme
} else {
p <- ggplot(mega_Data, aes_string(x = "x", y = "y")) +
facet_grid(xvar ~ yvar, scales = "free") +
geom_point(aes(colour = DependentVariable), na.rm = TRUE, alpha = alphaPoint, size = pointSize) +
stat_density(aes(x = x, y = ..scaled.. * diff(range(x)) + min(x)),
data = gg1$densities, position = "identity",
colour = "dodgerblue4", geom = "line", size = 1, alpha = 0.5) +
scale_color_manual(values = colours) +
guides(colour = guide_legend(override.aes = list(alpha = 1))) +
theme(panel.grid.minor = element_blank(), #remove gridlines
legend.position = "bottom", #legend at the bottom
axis.title.x = element_blank(), #remove x label
axis.title.y = element_blank() #remove y label
)#end theme
}
}
return (p)
}
#' Parallel Plot (Plot)
#'
#' Generate a plot of the columns of a data set for all or a range of instances. In
#' some cases this is useful to identify some patron.
#'
#' @param data an object of class "data.frame" containing just numerical columns.
#' @param rows an object of class "numeric" containing the list of rows
#' that you want in your parallel plot.
#' @param columns an object of class "numeric" containing the list of columns
#' that you want in your parallel plot.
#' @param dependentVariable an object of class "numeric", "factor" or "integer" is
#' a list of values containig the dependent variable.
#' @param dependentVariableName is an optional parameter. It's an string that
#' contains the name of your dependent variable.
#' @param lineSize is an optional parameter of class numeric with a single value
#' that represent the line size of plot.
#' @param alphaLine is an optional parameter of class numeric with a single value
#' that represent the alpha of lines in the plot.
#' @param x_lab a boolean that represent if you want or not the x axis scale. In
#' some cases, when you have many columns the plot could be ugly! The default value
#' is False.
#' @param colours is an optional parameter of class character with a list of colours
#' to use in the plot. The default value for continuos dependent variable is
#' c("darkred", "yellow", "darkgreen") and for categorical dependent variable are
#' the default colours defined by ggplot.
#'
#' @examples
#' #Example 1
#' iris.x <- iris[,1:4] # These are the independent variables
#' Species <- iris[,5] # This is the dependent variable
#'
#' # A ParallelPlot of all rows and all columns
#' ParallelPlot(iris.x, seq(1,nrow(iris.x),1), seq(1,ncol(iris.x),1), Species, "Species", 1, 0.5, TRUE)
#' # A ParallelPlot of all rows and some columns
#' ParallelPlot(iris.x, seq(1,nrow(iris.x),1), c(3,4), Species, "Species", 1, 0.5, TRUE)
#' # A ParallelPlot with a different colours palette
#' myPalette <- c("darkolivegreen4", "goldenrod1", "dodgerblue4")
#' ParallelPlot(iris.x, seq(1,nrow(iris.x),1), seq(1,ncol(iris.x),1), Species, "Species", 1, 0.5, TRUE, colours = myPalette)
#'
#'
#' #Example 2
#' # Getting a clean data set (without missing values)
#' cars <- read.csv("https://dl.dropboxusercontent.com/u/12599702/autosclean.csv", sep = ";", dec = ",")
#' cars.x <- cars[,1:16] # These are the independent variables
#' cars.y <- cars[,17] # This is the dependent variable
#'
#' # A ParallelPlot of all rows and all columns
#' ParallelPlot(cars.x, seq(1,nrow(cars.x),1), seq(1,ncol(cars.x),1), cars.y, "Price", 1, 0.5, TRUE)
#' # A ParallelPlot of all rows and some columns
#' ParallelPlot(cars.x, seq(1,nrow(cars.x),1), c(1,2,5,8,13,14), cars.y, "Price", 1, 0.8, TRUE)
#' # A ParallelPlot with a different colours palette
#' myPalette <- c("darkolivegreen4", "goldenrod1", "dodgerblue4")
#' ParallelPlot(cars.x, seq(1,nrow(cars.x),1), c(1,2,5,8,13,14), cars.y, "Price", 1, 0.8, TRUE, colours = myPalette)
ParallelPlot <- function(data, rows, columns, dependentVariable, dependentVariableName, lineSize, alphaLine, x_lab, colours) {
if (missing(data)) {
stop("Need to specify data!")
}
if (class(data) != "data.frame") {
stop("data must be a data.frame class!")
}
if (missing(rows)) {
stop("Need to specify rows!")
}
if (!(class(rows) == "numeric" || class(rows) == "integer")) {
stop("rows must be a numeric or integer class!")
}
if (missing(columns)) {
stop("Need to specify columns!")
}
if (!(class(columns) == "numeric" || class(columns) == "integer")) {
stop("columns must be a numeric or integer class!")
}
if (missing(dependentVariable)) {
stop("Need to specify dependentVariable!")
}
if (!(class(dependentVariable) == "numeric" || class(dependentVariable) == "factor" || class(dependentVariable) == "integer")) {
stop("dependentVariable must be a numeric, factor or integer class!")
}
if (missing(dependentVariableName)) {
dependentVariableName <- "Dependent Variable"
}
if (class(dependentVariableName) != "character") {
stop("dependentVariableName must be a character class!")
}
if (missing(lineSize)) {
lineSize <- 1
}
if (class(lineSize) != "numeric") {
stop("lineSize must be a numeric class!")
}
if (missing(alphaLine)) {
alphaLine <- 0.9
}
if (class(alphaLine) != "numeric") {
stop("alphaLine must be a numeric class!")
}
if (missing(x_lab)) {
x_lab = FALSE
}
if(class(x_lab) != "logical") {
stop("x_lab must be a logical class!")
}
if (missing(colours)) {
if (class(dependentVariable) == "numeric" || class(dependentVariable) == "integer") {
colours <- c("darkred", "yellow", "darkgreen")
}
}
#All parameters are OK!
x_name = "Wavelength"
subData <- data[rows,columns]
dependentVariable <- dependentVariable[rows]
rowsNum <- nrow(subData)
x <- seq(from = 1, to = rowsNum, length.out = rowsNum)
data <- data.frame(x, subData)
dataPlot <- melt(data, id = "x")
dataPlot <- data.frame(dataPlot, dependentVariable)
if (class(dependentVariable) == "numeric" || class(dependentVariable) == "integer") {
if (x_lab) {
min <- min(dependentVariable)
max <- max(dependentVariable)
p <- ggplot(dataPlot, aes(variable, value, group = x, colour = dependentVariable)) +
geom_line(size = lineSize, alpha = alphaLine) +
scale_color_gradientn(name = dependentVariableName,
colours = colours, breaks = c(min, max),
labels = c(min, max)) + #set the pallete
xlab(x_name) + ylab("Values") +
theme(legend.position = "bottom" #legend at the bottom
)#end theme
}
else {
min <- min(dependentVariable)
max <- max(dependentVariable)
p <- ggplot(dataPlot, aes(variable, value, group = x, colour = dependentVariable)) +
geom_line(size = lineSize, alpha = alphaLine) +
scale_color_gradientn(name = dependentVariableName,
colours = colours, breaks = c(min, max),
labels = c(min, max)) + #set the pallete
scale_x_discrete(breaks = c()) +
xlab(x_name) + ylab("Values") +
theme(legend.position = "bottom" #legend at the bottom
)#end theme
}
}
else {
if (x_lab) {
if (missing(colours)) {
p <- ggplot(dataPlot, aes(variable, value, group = x, colour = dependentVariable)) +
geom_line(size = lineSize, alpha = alphaLine) +
scale_color_discrete(name = dependentVariableName) +
xlab(x_name) + ylab("Values") +
guides(colour = guide_legend(override.aes = list(alpha = 1))) +
theme(legend.position = "bottom" #legend at the bottom
)#end theme +
} else {
p <- ggplot(dataPlot, aes(variable, value, group = x, colour = dependentVariable)) +
geom_line(size = lineSize, alpha = alphaLine) +
scale_color_manual(values = colours) +
xlab(x_name) + ylab("Values") +
guides(colour = guide_legend(override.aes = list(alpha = 1))) +
theme(legend.position = "bottom" #legend at the bottom
)#end theme
}
}
else {
if (missing(colours)) {
p <- ggplot(dataPlot, aes(variable, value, group = x, colour = dependentVariable)) +
geom_line(size = lineSize, alpha = alphaLine) +
scale_color_discrete(name = dependentVariableName) +
scale_x_discrete(breaks = c()) +
xlab(x_name) + ylab("Values") +
guides(colour = guide_legend(override.aes = list(alpha = 1))) +
theme(legend.position = "bottom" #legend at the bottom
)#end theme
} else {
p <- ggplot(dataPlot, aes(variable, value, group = x, colour = dependentVariable)) +
geom_line(size = lineSize, alpha = alphaLine) +
scale_color_manual(values = colours) +
scale_x_discrete(breaks = c()) +
xlab(x_name) + ylab("Values") +
guides(colour = guide_legend(override.aes = list(alpha = 1))) +
theme(legend.position = "bottom" #legend at the bottom
)#end theme
}
}
}
return (p)
}
#' Density Plot (Plot)
#'
#' Generate a density plot for a specific column of the data.
#'
#' @param data an object of class data frame with the data.
#' @param col an integer that specify the column that you want for make the plot.
#'
#' @seealso http://www.rdatamining.com/examples/outlier-detection
#'
#' @examples
#' #Example 1
#' #install.packages("Rlof")
#' library(Rlof) #for outlier detection
#'
#' iris.x <- iris[,1:4] # These are the independent variables
#' Species <- iris[,5] # This is the dependent variable
#'
#' DensityPlot(iris.x,1)
#'
#'
#' #Example 2
#' #install.packages("Rlof")
#' library(Rlof) #Outlier detection library
#'
#' iris.x <- iris[,1:4] # These are the independent variables
#' Species <- iris[,5] # This is the dependent variable
#'
#' outlier.scores <- lof(iris.x, k = 5) #applying outlier detection
#' outlier.scores <- data.frame(outlier.scores)
#' DensityPlot(outlier.scores, 1) #Generating a plot of outliers scores
#'
#'
#' #Example 3
#' #install.packages("Rlof")
#' library(Rlof) #Outlier detection library
#'
#' iris.x <- iris[,1:4] # These are the independent variables
#' Species <- iris[,5] # This is the dependent variable
#'
#' outlier.scores <- lof(iris.x, k = c(5:10)) #applying outlier detection
#' mean <- rowMeans(outlier.scores) #Calculating the mean of every execution
#' outlier.scores <- data.frame(outlier.scores, mean) #adding mean to data frame
#' DensityPlot(outlier.scores, ncol(outlier.scores)) #Generating a plot of outliers scores
#'
#'
#' #Example 4
#' #install.packages("Rlof")
#' library(Rlof) #Outlier detection library
#' library(plyr)
#' # Getting a data set without missing values
#' cars <- read.csv("https://dl.dropboxusercontent.com/u/12599702/autosclean.csv", sep = ";", dec = ",")
#'
#' cars.x <- cars[,1:16] # These are the independent variables
#' cars.y <- cars[,17] # This is the dependent variable
#'
#' outlier.scores <- lof(cars.x, k = c(5:10)) #applying outlier detection
#' mean <- rowMeans(outlier.scores) #Calculating the mean of every execution
#' outlier.scores <- data.frame(outlier.scores, mean) #adding mean to data frame
#' DensityPlot(outlier.scores, ncol(outlier.scores)) #Generating a plot of outliers scores
#'
#' aux <- outlier.scores[,7]>1.7 #1.7 is the threshold selected
#' count(aux)[2,2] #Number of outliers found
#' outliers <- order(outlier.scores[,7], decreasing=T)[1:count(aux)[2,2]] #Getting the values that are on the threshold
#' Score <- outlier.scores[outliers,7] #Getting outliers scores
#' outliers <- data.frame(outliers,Score)
#' names(outliers) <- c("Position","Score")
#' View(outliers)
#'
#' auxOutliers <- outlier.scores[-outliers[1:3,1],] #Eliminating the 3 most remote instances!
#' DensityPlot(auxOutliers, ncol(outlier.scores)) #Generating a plot of outliers scores
DensityPlot <- function(data, col) {
if (missing(data)) {
stop("Need to specify data!")
}
if (class(data) != "data.frame") {
stop("data must be a data frame class!")
}
if (missing(col)) {
stop("Need to specify col!")
}
if (!(class(col) == "numeric" || class(col) == "integer")) {
stop("col must be a numeric or integer class!")
}
if (col > ncol(data)) {
stop("Col value must be less than ncol of data!")
}
#All parameters are OK!
names(data)[col] <- "mean"
p <- ggplot(data, aes(x = mean)) +
geom_density(colour="darkgreen", fill="darkgreen", alpha=0.3) +
xlab("Values") +
ylab("Density")
return (p)
}
#' Plot PC 3D (3DPlot)
#'
#' Generate a 3D plot Generates a 3D graphic for a set of 3 columns of the data set.
#'
#' @param data an object of class data frame with the data.
#' @param columns an object of class "numeric" containing the list of columns
#' that you want in your parallel plot.
#' @param dependentVariable an object of class "numeric", "factor" or "integer" is
#' a list of values containig the dependent variable.
#'
#' @examples
#' #Example 1
#' iris.x <- iris[,1:4] # These are the independent variables
#' Species <- iris[,5] # This is the dependent variable
#'
#' # 3D Plot of 3 first columns of data set
#' Plot3D(iris.x, c(1,2,3), Species)
#'
#'
#' #Example 2
#' iris.x <- iris[,1:4] # These are the independent variables
#' Species <- iris[,5] # This is the dependent variable
#'
#' ir.pca <- prcomp(iris.x, center = TRUE, scale. = TRUE) # performin prcomp
#'
#' # 3D Plot of 3 first columns of data set
#' Plot3D(as.data.frame(ir.pca$x), c(1,2,3), Species)
#'
#'
#' #Example 3
#' # Getting a clean data set (without missing values)
#' cars <- read.csv("https://dl.dropboxusercontent.com/u/12599702/autosclean.csv", sep = ";", dec = ",")
#' cars.x <- cars[,1:16] # These are the independent variables
#' cars.y <- cars[,17] # This is the dependent variable
#'
#' # 3D Plot of 3 first columns of data set
#' Plot3D(cars.x, c(1,2,3), cars.y)
#'
#'
#' #Example 4
#' # Getting a clean data set (without missing values)
#' cars <- read.csv("https://dl.dropboxusercontent.com/u/12599702/autosclean.csv", sep = ";", dec = ",")
#' cars.x <- cars[,1:16] # These are the independent variables
#' cars.y <- cars[,17] # This is the dependent variable
#'
#' cars.pca <- prcomp(cars.x, center = TRUE, scale. = TRUE) # performin prcomp
#'
#' # 3D Plot of 3 first columns of data set
#' Plot3D(as.data.frame(cars.pca$x), c(1,2,3), cars.y)
Plot3D <- function(data, columns, dependentVariable){
if (missing(data)) {
stop("Need to specify data!")
}
if (class(data) != "data.frame") {
stop("data must be a data.frame class!")
}
if (missing(columns)) {
stop("Need to specify columns!")
}
if (!(class(columns) == "numeric" || class(columns) == "integer")) {
stop("columns must be a numeric or integer class!")
}
if (length(columns) != 3) {
stop("The number of selected columns must be 3!")
}
if (missing(dependentVariable)) {
stop("Need to specify dependentVariable!")
}
if (!(class(dependentVariable) == "numeric" || class(dependentVariable) == "factor" || class(dependentVariable) == "integer")) {
stop("dependentVariable must be a numeric, factor or integer class!")
}
#All parameters are OK!
subData<-data[,columns]
col1 <- subData[,1]
col2 <- subData[,2]
col3 <- subData[,3]
x_lab <- colnames(subData)[1]
y_lab <- colnames(subData)[2]
z_lab <- colnames(subData)[3]
if (class(dependentVariable) == "numeric" || class(dependentVariable) == "integer") {
cols <- myColorRamp(c("darkred", "yellow", "darkgreen"), dependentVariable)
plot3d(x = col1, y = col2, z = col3, col = cols, size = "4", xlab = x_lab, ylab = y_lab, zlab = z_lab)
}
else {
cols <- myColorRamp(c("darkred", "yellow", "darkgreen"), as.numeric(dependentVariable))
plot3d(x = col1, y = col2, z = col3, col = cols, size = "4", xlab = x_lab, ylab = y_lab, zlab = z_lab)
}
}
#' Simple Plot of 2 columns (Plot)
#'
#' Generate a plot of 2 columns of data set using ggplot. You must indicate which
#' columns you want in the graph.
#'
#' @param data an object of class data frame with the data.
#' @param DependentVariable an object of class "numeric", "factor" or "integer" is
#' a list of values containig the dependent variable.
#' @param x_axis an integer that represent the number of the column that you want
#' in your x axis.
#' @param y_axis an integer that represent the number of the column that you want
#' in your y axis.
#' @param dependentVariableName is an optional parameter. It's an string that
#' contains the name of your dependent variable.
#' @param pointSize is an optional parameter of class numeric with a single value
#' that represent the point size of plot.
#' @param alphaPoint is an optional parameter of class numeric with a single value
#' that represent the alpha of points in the plot.
#' @param colours is an optional parameter of class character with a list of colours
#' to use in the plot. The default value for continuos dependent variable is
#' c("darkred", "yellow", "darkgreen") and for categorical dependent variable are
#' the default colours defined by ggplot.
#'
#' @seealso elbowPlot
#'
#' @examples
#' #Example 1
#' iris.x <- iris[,1:4] # These are the independent variables
#' Species <- iris[,5] # This is the dependent variable
#'
#' # Plot of first 2 columns of data set
#' simplePlot(iris.x, Species, 1, 2, "Species", 2, 0.9)
#' # A plot with a different colours palette
#' myPalette <- c("darkolivegreen4", "goldenrod1", "dodgerblue4")
#' simplePlot(iris.x, Species, 1, 2, "Species", 2, 0.9, colours = myPalette)
#'
#'
#' #Example 2
#' iris.x <- iris[,1:4] # These are the independent variables
#' Species <- iris[,5] # This is the dependent variable
#'
#' ir.pca <- prcomp(iris.x, center = TRUE, scale. = TRUE) #performing prcomp
#'
#' # Plot of first 2 columns of principal components
#' simplePlot(as.data.frame(ir.pca$x), Species, 1, 2, "Species", 2, 0.9)
#' # A plot with a different colours palette
#' myPalette <- c("darkolivegreen4", "goldenrod1", "dodgerblue4")
#' simplePlot(as.data.frame(ir.pca$x), Species, 1, 2, "Species", 2, 0.9, colours = myPalette)
#'
#'
#' #Example 3
#' # Getting a clean data set (without missing values)
#' cars <- read.csv("https://dl.dropboxusercontent.com/u/12599702/autosclean.csv", sep = ";", dec = ",")
#' cars.x <- cars[,1:16] # These are the independent variables
#' cars.y <- cars[,17] # This is the dependent variable
#'
#' cars.pca <- prcomp(cars.x, center = TRUE, scale. = TRUE) #performing prcomp
#'
#' # Plot of first 2 columns of principal components
#' simplePlot(as.data.frame(cars.pca$x), cars.y, 1, 2, "Price", 2, 0.9)
#' # A plot with a different colours palette
#' myPalette <- c("darkolivegreen4", "goldenrod1", "dodgerblue4")
#' simplePlot(as.data.frame(cars.pca$x), cars.y, 1, 2, "Price", 2, 0.9, colours = myPalette)
simplePlot <- function(data, DependentVariable, x_axis, y_axis, dependentVariableName, pointSize, alphaPoint, colours) {
if (missing(data)) {
stop("Need to specify data!")
}
if (class(data) != "data.frame") {
stop("data must be a data.frame class!")
}
if (missing(DependentVariable)) {
stop("Need to specify DependentVariable!")
}
if (!(class(DependentVariable) == "numeric" || class(DependentVariable) == "factor" || class(DependentVariable) == "integer")) {
stop("DependentVariable must be a numeric, factor or integer class!")
}
if (missing("x_axis")) {
stop("Need to specify x_axis!")
}
if (!(class(x_axis) == "numeric" || class(x_axis) == "integer")) {
stop("x_axis must be a numeric or integer class!")
}
if (missing("y_axis")) {
stop("Need to specify y_axis!")
}
if (!(class(y_axis) == "numeric" || class(y_axis) == "integer")) {
stop("y_axis must be a numeric or integer class!")
}
if (missing(dependentVariableName)) {
dependentVariableName <- "Dependent Variable"
}
if (class(dependentVariableName) != "character") {
stop("dependentVariableName must be a character class!")
}
if (missing(pointSize)) {
pointSize <- 1
}
if (class(pointSize) != "numeric") {
stop("pointSize must be a numeric class!")
}
if (missing(alphaPoint)) {
alphaPoint <- 0.5
}
if (class(alphaPoint) != "numeric") {
stop("alphaPoint must be a numeric class!")
}
if (missing(colours)) {
if (class(DependentVariable) == "numeric" || class(DependentVariable) == "integer") {
colours <- c("darkred", "yellow", "darkgreen")
}
}
#All parameters are OK!
subData <- data.frame(data[,x_axis], data[,y_axis], DependentVariable)
x_axis <- colnames(data)[x_axis]
y_axis <- colnames(data)[y_axis]
names(subData) <- c(x_axis, y_axis, "DependentVariable")
if (class(DependentVariable) == "numeric" || class(DependentVariable) == "integer") {
min <- min(DependentVariable)
max <- max(DependentVariable)
p <- ggplot(subData, aes_string(x = x_axis, y = y_axis)) +
geom_point(aes(colour = DependentVariable), na.rm = TRUE, alpha = alphaPoint, size = pointSize) +
scale_color_gradientn(name = dependentVariableName,
colours = colours, breaks = c(min, max),
labels = c(min, max)) + #set the pallete
theme(panel.grid.minor = element_blank(), #remove gridlines
legend.position = "bottom" #legend at the bottom
) + #end theme
xlab(x_axis) + ylab(y_axis)
}
else {
if (missing(colours)) {
p <- ggplot(subData, aes_string(x = x_axis, y = y_axis)) +
geom_point(aes(colour = DependentVariable), na.rm = TRUE, alpha = alphaPoint, size = pointSize) +
scale_color_discrete(name = dependentVariableName) +
guides(colour = guide_legend(override.aes = list(alpha = 1))) +
theme(panel.grid.minor = element_blank(), #remove gridlines
legend.position = "bottom" #legend at the bottom
) + #end theme
xlab(x_axis) + ylab(y_axis)
} else {
p <- ggplot(subData, aes_string(x = x_axis, y = y_axis)) +
geom_point(aes(colour = DependentVariable), na.rm = TRUE, alpha = alphaPoint, size = pointSize) +
scale_color_manual(values = colours) +
guides(colour = guide_legend(override.aes = list(alpha = 1))) +
theme(panel.grid.minor = element_blank(), #remove gridlines
legend.position = "bottom" #legend at the bottom
) + #end theme
xlab(x_axis) + ylab(y_axis)
}
}
return (p)
}
#' Ridge Plot
#'
#' Plot the cross-validation curve produced by cv.glmnet using ggplot
#'
#' @param ridgeModelCV an object of class cv.glmnet that contain the cross-validation
#' information for generate a curve.
#' @param pointSize is an optional parameter of class numeric with a single value
#' that represent the point size of plot. The default value is 1.5
#' @param alphaPoint is an optional parameter of class numeric with a single value
#' that represent the alpha of points in the plot. The default value is 0.8
#' @param colourPoint is an optional parameter of class character with a single
#' colour. The default value is "darkred".
#' @param colourLine is an optional parameter of class character with a single
#' colour. The default value is "dodgerblue4".
#' @param errorMode is an optional parameter of class character that defined if the
#' error will be displayed using "errorbar" or "ribbon". The default value is "errorbar"
#' @param colourError is an optional parameter of class character with a single
#' colour. The default value is "#A9A9A9".
#'
#' @examples
#' #Example 1
#' #install.packages("MASS")
#' library(MASS) #for ridge regression
#' #install.packages("glmnet")
#' library(glmnet) #for parameter optimization
#' iris.x <- iris[,1:3] # These are the independent variables
#' iris.y <- iris[,4] # This is the dependent variable
#'
#' #We perform the ridge method
#' grid <- 10^seq(10, -2, length = 100) #Define a large grid for lambda values
#' set.seed(2015)
#' ridge <- cv.glmnet(as.matrix(iris.x), iris.y, alpha = 0, lambda = grid) #alpha = 0 for ridge, alpha = 1 for lasso
#' RidgePlot(ridge)
#'
#' bestLambda <- ridge$lambda.min # The optimal lambda
#' ridge.final <- lm.ridge(y ~ ., data = X, lambda = bestLambda)
#'
#'
#' #Example 2
#' #install.packages("MASS")
#' library(MASS) #for ridge regression
#' #install.packages("glmnet")
#' library(glmnet) #for parameter optimization
#' # Getting a clean data set (without missing values)
#' cars <- read.csv("https://dl.dropboxusercontent.com/u/12599702/autosclean.csv", sep = ";", dec = ",")
#' cars.x <- cars[,1:16] # These are the independent variables
#' cars.y <- cars[,17] # This is the dependent variable
#'
#' #We perform the ridge method
#' grid <- 10^seq(10, -2, length = 100) #Define a large grid for lambda values
#' set.seed(2015)
#' ridge <- cv.glmnet(as.matrix(cars.x), cars.y, alpha = 0, lambda = grid) #alpha = 0 for ridge, alpha = 1 for lasso
#' RidgePlot(ridge)
#' RidgePlot(ridge, errorMode = "ribbon") #We change the plot with geom_ribbon for errors
#'
#' bestLambda <- ridge$lambda.min # The optimal lambda
#' ridge.final <- lm.ridge(y ~ ., data = X, lambda = bestLambda)
RidgePlot <- function(ridgeModelCV, pointSize, alphaPoint, colourPoint, colourLine, errorMode, colourError) {
if (missing(ridgeModelCV)) {
stop("Need to specify ridgeModelCV!")
}
if (class(ridgeModelCV) != "cv.glmnet") {
stop("ridgeModelCV must be a cv.glmnet class!")
}
if (missing(pointSize)) {
pointSize <- 1.5
}
if (class(pointSize) != "numeric") {
stop("pointSize must be a numeric class!")
}
if (missing(alphaPoint)) {
alphaPoint <- 0.8
}
if (class(alphaPoint) != "numeric") {
stop("alphaPoint must be a numeric class!")
}
if (missing(colourPoint)) {
colourPoint <- "darkred"
}
if (class(colourPoint) != "character") {
stop("colourPoint must be a character class!")
}
if (missing(colourLine)) {
colourLine <- "dodgerblue4"
}
if (class(colourLine) != "character") {
stop("colourLine must be a character class!")
}
if (missing(errorMode)) {
errorMode <- "errorbar"
}
if (class(errorMode) != "character") {
stop("errorMode must be a character class!")
} else {
if (!(errorMode== "errorbar" || errorMode == "ribbon")) {
stop("errorMode must be a errorbar or ribbon!")
}
}
if (missing(colourError)) {
colourError <- "#A9A9A9"
}
if (class(colourError) != "character") {
stop("colourError must be a character class!")
}
tidied_cv <- tidy(ridgeModelCV)
glance_cv <- glance(ridgeModelCV)
tidied_cv <- data.frame(tidied_cv, log(tidied_cv$lambda))
names(tidied_cv) <- c(names(tidied_cv)[1:6], "logLambda")
if (errorMode == "errorbar") {
p <- ggplot(tidied_cv, aes(logLambda, estimate)) +
xlab("log(Lambda)") + ylab("MSE") +
geom_errorbar(aes(ymin = conf.low, ymax = conf.high), colour = colourError, width = .5) +
geom_vline(xintercept = log(glance_cv$lambda.min), lty = 2, colour = colourLine) +
geom_vline(xintercept = log(glance_cv$lambda.1se), lty = 2, colour = colourLine) +
geom_point(colour = colourPoint, size = pointSize, alpha = alphaPoint)
} else {
p <- ggplot(tidied_cv, aes(logLambda, estimate)) +
xlab("log(Lambda)") + ylab("MSE") +
geom_ribbon(aes(ymin = conf.low, ymax = conf.high), alpha = .25, colour = colourError) +
geom_vline(xintercept = log(glance_cv$lambda.min), lty = 2, colour = colourLine) +
geom_vline(xintercept = log(glance_cv$lambda.1se), lty = 2, colour = colourLine) +
geom_point(colour = colourPoint, size = pointSize, alpha = alphaPoint)
}
return (p)
}
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Embedding an R snippet on your website
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.