comp.dist.plot: Plotting Two Distributions Side by Side for Comparison
In lindayi/distcomp: Plotting Distribution Differences

Description Usage Arguments Value Author(s) References Examples

View source: R/comp.dist.plot.R

Plotting both distributions in one beanplot, and calculate the significance of difference between two distributions using Wilcoxon Rank Sum or Signed Rank Test. The verticle line shows the median values of the distributions. Use log = "x" to force logging the x-axis.

comp.dist.plot(dist1, dist2,
               legend1 = "Distribution 1",
               legend2 = "Distribution 2",
               legendpos = "topright",
               col1 = "#C8C8C8", col2 = "#646464",
               xlab = "", cut = TRUE,
               paired = FALSE,
               ...)

`dist1`	the first distribution for comparison
`dist2`	the second distribution for comparison
`legend1`	the label for the legend of the first distribution
`legend2`	the label for the legend of the second distribution
`legendpos`	the position of the legend, specified by a single keyword from the list "bottomright", "bottom", "bottomleft", "left", "topleft", "top", "topright", "right" and "center".
`col1`	the color for the first distribution
`col2`	the color for the second distribution
`xlab`	the label for the x-axis
`cut`	logical; if TRUE, the beanplot will be cut at the minimum / maximum value of the distribution to avoid extending the beanplot to illegal values
`paired`	logical; if TRUE, a Wilcoxon signed rank test of the null that `dist1` - `dist2` is symmetric about 0 (or `mu` if specified) is performed; if FALSE, a Wilcoxon rank sum test (equivalent to the Mann-Whitney test: see the Note) is carried out. In this case, the null hypothesis is that `dist1` and `dist2` differ by a location shift of 0 (or `mu` if specified) and the alternative is that they differ by some other location shift. See `wilcox.test` for details
`...`	further arguments to be passed to `wilcox.test`

A list of class htest. See wilcox.test for details.

Maintainer: Dayi Lin http://lindayi.me

David F. Bauer (1972), Constructing confidence sets using rank statistics. Journal of the American Statistical Association 67, 687<e2><80><93>690.

Please consider citing the following paper when using this package and the visualization style of effect size:

Lin, D., Bezemer, C. P., & Hassan, A. E. (2018). An empirical study of early access games on the Steam platform. Empirical Software Engineering, 23(2), 771-799.

temp.dist1 = rnorm(100,1,2)
temp.dist2 = rnorm(100,-1,2)

# Basic (by default unpaired)
comp.dist.plot(temp.dist1, temp.dist2)

# Without cut
comp.dist.plot(temp.dist1, temp.dist2, cut = FALSE)

# Paired test
comp.dist.plot(temp.dist1, temp.dist2, paired = TRUE)