Permalink
Cannot retrieve contributors at this time
| # | |
| # grant-sackman.R, 4 Mar 20 | |
| # Data from: | |
| # The 28:1 {Grant}/{Sackman} legend is misleading, or: {How} large is interpersonal variation really? | |
| # Lutz Prechelt | |
| # | |
| # Example from: | |
| # Evidence-based Software Engineering: based on the publicly available data | |
| # Derek M. Jones | |
| # | |
| # TAG experiment_human developer_performance | |
| source("ESEUR_config.r") | |
| library("vioplot") | |
| pal_col=rainbow(8) | |
| gs=read.csv(paste0(ESEUR_dir, "developers/grant-sackman.csv.xz"), as.is=TRUE) | |
| vioplot(time ~ group+task, data=gs, horizontal=TRUE, col=pal_col, | |
| xaxs="i", | |
| ylim=c(0, max(gs$time)), | |
| xlab="Time (minutes)", ylab="") | |
| # library("plyr") | |
| # | |
| # daply(gs, .(task), function(df) max(df$time)/min(df$time)) | |
| # daply(gs, .(task, group), function(df) max(df$time)/min(df$time)) | |