#

# 1611-09187a.R, 8 Nov 19

# Data from:

# Benjamin Danglot and Philippe Preux and Benoit Baudry and Martin Monperrus

# Correctness Attraction: {A} Study of Stability of Software Behavior Under Runtime Perturbation

#

# Example from:

# Evidence-based Software Engineering: based on the publicly available data

# Derek M. Jones

#

# TAG runtime_perturbation

source("ESEUR_config.r")

library("plyr")

library("vioplot")

normalise_pert=function(df)

{

ao=subset(df, Perturb_str == "AddOne")

# Percentage of success, as an integer

perc_succ=trunc(100*ao$Success/ao$Num_Perturb)

perc_succ[is.nan(perc_succ)]=0

perc_succ[is.infinite(perc_succ)]=0

# Weight location by percentage successes

NormLoc=rep(100*ao$IndexLoc/max(ao$IndexLoc),

times=perc_succ)

return(data.frame(Program=df$Program[1], NormLoc))

}

ptb=read.csv(paste0(ESEUR_dir, "reliability/1611-09187a.csv.xz"), as.is=TRUE)

subjects=c("quicksort", "zip", "sudoku", "md5", "rsa", "rc4", "canny", "lcs", "laguerre", "linreg")

ptb=subset(ptb, Program %in% subjects)

vp=ddply(ptb, .(Program), normalise_pert)

pal_col=rainbow(length(unique(vp$Program)))

vioplot(NormLoc ~ Program, data=vp,

h=1.1,

horizontal=TRUE,

col=pal_col, border=pal_col,

xaxs="i",

xlab="Perturbation location (normalised)", ylab="")