ascknd

Using the DoE package of R, I create the following experimental design:
Design.1 <- fac.design(nfactors= 5 ,replications= 20 ,repeat.only= FALSE ,blocks= 1 ,
  randomize= TRUE ,seed= 25027 ,nlevels=c( 2,2,2,2,2 ), factor.names=list( read=c(0,100),
  rand=c(0,100),blk_sz=c(4,8),protocol=c(‘”fc”‘,'”nfs”‘),iops=c(1000,4000) ) )

The design can then be exported for later use:
export.design(Design.1, type=”all”,path=”V:/Doe”, file=”Design.1″,
  replace=FALSE)

I use this design to analyse the issues raised in this post:

NFS vs Fibre Channel: Comparing CPU Utilization in VMWare

here:

Analysing Experimental Results With R

Tagged: fibre channel, NFS, performance, R

§ 2 Responses to Creating an Experimental Design in R

• Analysing Experimental Results With R | ascknd says:

  May 27, 2014 at 11:59 am

  […] ← Creating an Experimental Design in R […]

  Reply
• NFS vs Fibre Channel: Comparing CPU Utilization in VMWare | ascknd says:

  May 27, 2014 at 12:13 pm

  […] https://ascknd.com/2014/05/27/creating-an-experimental-design-in-r/ […]

  Reply