Skip to content
Permalink
master
Switch branches/tags
Go to file
 
 
Cannot retrieve contributors at this time
#
# TUD-SERG-2016-004.R, 18 Mar 17
# Data from:
# Do Estimators Learn? {On} the Effect of a Positively Skewed Distribution of Effort Data on Software Portfolio Productivity
# Hennie Huijgens and Frank Vogelezang
#
# Example from:
# Evidence-based Software Engineering: based on the publicly available data
# Derek M. Jones
#
# TAG estimate_learn estimate_release estimate_actual
source("ESEUR_config.r")
pal_col=rainbow(2)
est=read.csv(paste0(ESEUR_dir, "economics/TUD-SERG-2016-004.csv.xz"), as.is=TRUE)
est$seq=1:nrow(est)
major_rel=subset(est, trunc(Release) == Release)
minor_rel=subset(est, trunc(Release) != Release)
plot(est$seq, est$Actual_Effort/est$Planned_Effort, type="l", col="grey",
xlim=c(1, nrow(est)), ylim=c(min(est$Actual_Effort/est$Planned_Effort),
max(est$Actual_Effort/est$Planned_Effort)),
xlab="Release", ylab="Actual/Planned effort\n")
points(major_rel$seq, major_rel$Actual_Effort/major_rel$Planned_Effort, col=pal_col[1])
points(minor_rel$seq, minor_rel$Actual_Effort/minor_rel$Planned_Effort, col=pal_col[2])
legend(x="topright", legend=c("Major releases", "Point releases"), bty="n", fill=pal_col, cex=1.2)