violinplotter {violinplotter}

Plotting and Comparing Means with Violin Plots

Description

Plotting and Comparing Means with Violin Plots

Usage

violinplotter(formula, 
                     data=NULL, 
                     TITLE="", 
                     XLAB="", 
                     YLAB="", 
                     VIOLIN_COLOURS=c("#e0f3db","#a8ddb5","#7bccc4","#2b8cbe"), 
                     PLOT_BARS=TRUE, 
                     ERROR_BAR_COLOURS=c("#636363","#1c9099","#de2d26"), 
                     SHOW_SAMPLE_SIZE=FALSE, 
                     SHOW_MEANS=TRUE, 
                     CATEGORICAL=TRUE, 
                     LOGX=FALSE, 
                     LOGX_BASE=10, 
                     MANN_WHITNEY=TRUE, 
                     HSD=FALSE, 
                     ALPHA=0.05, 
                     REGRESS=FALSE)

Arguments

formula

R's compact symbolic form to represent linear models with fixed additive and interaction effects (See ?formula for more information) [mandatory]

data

data.frame containing the response and explanatory variables which forms the formula above [default=NULL]

TITLE

string or vector of strings corresponding to violin plot title/s [default: combinations of the "response variable name X explanatory variable" from the dataframe column names]

XLAB

string or vector of strings specifying the x-axis labels [default: column names of the explanatory variables (and their combinations) from data]

YLAB

string or vector of strings specifying the y-axis labels [default: column names of the response variable from data]

VIOLIN_COLOURS

vector or list of vectors of colors of the violin plots which are repeated if the length is less than the number of explanatory factor levels or less than the number of explanatory factors in the case of a list [default=c("#e0f3db", "#ccebc5", "#a8ddb5", "#7bccc4", "#4eb3d3", "#2b8cbe")]

PLOT_BARS

logical (i.e. TRUE or FALSE) to plot all or none of the bars; or vector strings which bars to plot (e.g. "stdev", "sterr", "ci") [default=TRUE=c("stdev", "sterr", "ci")]

ERROR_BAR_COLOURS

vector of colors of standard deviation, standard error and 95 percent confidence interval error bars (error bar selection via leaving one of the three colors empty) [default=c("#636363", "#1c9099", "#de2d26")]

SHOW_SAMPLE_SIZE

logical referring to whether or not to show the sample sizes for each category [default=FALSE]

SHOW_MEANS

logical referring to whether or not to show the means [default=TRUE]

CATEGORICAL

logical or vector of logicals referring to whether the explanatory variable/s is/are strictly categorical [default=TRUE]

LOGX

logical or vector of logicals referring to whether to transform the explanatory variable/s into the logarithm scale [default=FALSE]

LOGX_BASE

numeric or vector of numerics referring to the logarithm base to transform the explanatory variable/s with [default=1]

MANN_WHITNEY

logical or vector of logicals referring to whether to perform Mann-Whitney Grouping [default=TRUE]

HSD

logical or vector of logicals referring to whether to perform Tukey's Honest Significance Grouping [default=FALSE]

ALPHA

numeric significance level for the analysis of variance F-test and Tukey's mean comparison [default=0.05]

REGRESS

logical or vector of logicals referring to whether to regress the response variable against the explanatory variable/s [default=FALSE]

Value

Violin plot/s with optional error bars, mean comparison grouping/s, and regression line/s

Mean comparison grouping/s based on Tukey's Hones significant difference and regression line statistics, if applicable

Examples

x1 = rep(rep(rep(c(1:5), each=5), times=5), times=5)
x2 = rep(rep(letters[6:10], each=5*5), times=5)
x3 = rep(letters[11:15], each=5*5*5)
y = rep(1:5, each=5*5*5) + rnorm(rep(1:5, each=5), length(x1))
formula = log(y) ~ exp(x1) + x2 + x3 + (x2:x3)
test1 = violinplotter(formula=formula)
test2 = violinplotter(formula=formula, PLOT_BARS=c("ci", "stdev"))