BoxPlot: Boxplot with confidence interval and ANOVA on the plot.

In CorReg: Linear Regression Based on Linear Structure Between Variables

Description

classical boxplot function improved with integrated confidence level on the mean for each group ploted on the graph and also ANOVA with p-value and its interpretation given in the legend.

Usage

BoxPlot(
  num,
  grp = NULL,
  data = NULL,
  AnoVa = TRUE,
  risk = 0.05,
  lang = c("en", "fr"),
  verbose = TRUE,
  ...
)

Arguments

num	a numeric vector to plot boxplot(num~grp). Represents the value that will be compared between the groups.
grp	a qualitative vector (factor) to plot boxplot(num~grp). Represents the groups we will compare.
data	a data.frame (or list) from which the variables in formula should be taken.
AnoVa	boolean to compute or not anova (when multiple groups) to see if they differ in mean.If false the Kruskal-Wallis Rank Sum test is computed instead.
risk	the risk value used for confidence intervals.
lang	lingustic parameter to specify the language of the legend
verbose	boolean to make a test and print the result in the subtitle
...	Other graphical parameters

Examples

repart = c(20, 40, 40)
X = data.frame(num = c(rnorm(repart[1], 10, 1), rnorm(repart[2], 11, 1), rnorm(repart[3], 10, 1)),
               grp = rep(c("A", "B", "C"), repart))


BoxPlot(X$num, X$grp, data = X, ylab = "num", main = "boxplot with confidence intervals")
# Confidence interval in red with mean in blue.

CorReg documentation built on Feb. 20, 2020, 5:07 p.m.