BookmarkSubscribeRSS Feed
IanWakeling
Barite | Level 11

I thought there might be some users here who would like to see how the hadamard function can be used to generate balanced incomplete block designs in IML.

The code attached implements combinatorial constructions for the designs rather than taking the iterative search approach that is possible with PROC OPTEX.

2 REPLIES 2
Rick_SAS
SAS Super FREQ

Thanks for posting. I don't know much about this topic, but people who are interested in generating experimental designs should look at the collection of macros that use SAS/STAT, SAS/QC, and SAS/IML to generate all kinds of experimental designs, including ones that use Hadamard matrices. If you are interested in these topics, see http://support.sas.com/resources/papers/tnote/tnote_marketresearch.html, which includes a link to the MktBIBD macro, and lots more!

IanWakeling
Barite | Level 11

Thanks for posting the link Rick.  I think the market research macros are using Hadamard matrices as a source of orthogonal arrays and fractional factorials, not to generate BIBDs.  So while MktBIBD is a good tool for finding partially balanced designs, it will generally be quite bad at finding these Hadamard designs as it also uses the OPTEX approach.  The point at which things start to break is around 19 samples,  so

%MktBIBD(b=19,t=19,k=9,options=);

will find the balanced design with about 4 minutes CPU on my computer, while

print (sbibd(19));

with the IML program is virtually instantaneous.  Any higher than 19 and you would have to be content with a sub-optimal design if you use MktBIBD.

 

sas-innovate-white.png

Our biggest data and AI event of the year.

Don’t miss the livestream kicking off May 7. It’s free. It’s easy. And it’s the best seat in the house.

Join us virtually with our complimentary SAS Innovate Digital Pass. Watch live or on-demand in multiple languages, with translations available to help you get the most out of every session.

 

Register now!

From The DO Loop
Want more? Visit our blog for more articles like these.
Discussion stats
  • 2 replies
  • 1298 views
  • 0 likes
  • 2 in conversation