Re: [R] A combinatorial assignment problem

2014-05-01 Thread Charles Berry
Ravi Varadhan jhu.edu> writes: > > Thanks, Bert. > I have written this simple code, which is crude, but seems to do a decent job. It works perfectly when M is a > factor of R. Otherwise, it gives decent balance (of course, balance is not guaranteed). I guess it is > possible to take the res

Re: [R] A combinatorial assignment problem

2014-05-01 Thread Ravi Varadhan
Thank you, Dan and Bert. Bert - Your approach provides a solution. However, it has the undesired property of referees lumping together (I apologize that I did not state this as a condition). In other words, it does not "mix" the referees in some random fashion. Dan - your approach attempts

Re: [R] A combinatorial assignment problem

2014-05-01 Thread Ravi Varadhan
Thanks, Bert. I have written this simple code, which is crude, but seems to do a decent job. It works perfectly when M is a factor of R. Otherwise, it gives decent balance (of course, balance is not guaranteed). I guess it is possible to take the results, that are somewhat unbalanced and th

Re: [R] A combinatorial assignment problem

2014-05-01 Thread Charles Berry
Ravi Varadhan jhu.edu> writes: > > Hi, > > I have this problem: K candidates apply for a job. There are R referees available to review their resumes and > provide feedback. Suppose that we would like M referees to review each candidate (M < R). How would I assign > candidates to referees (o

Re: [R] A combinatorial assignment problem

2014-05-01 Thread Bert Gunter
Ravi: You cannot simultaneously have balance and guarantee random mixing. That is, you would need to specify precisely what you mean by balance and random mixing in this context, as these terms are now subjective and undefined. You could, of course, randomize the initial assignment of referees to

Re: [R] A combinatorial assignment problem

2014-05-01 Thread Bert Gunter
I had trouble with my email and it went before it should. Here's the solution I meant to send: Arrange the r referees in a circle. start <- 0 Replicate k times{ end <- (start + m-1)%% r output: c(start,end) +1 start <- (end+1)%% r } The start and end pairs give the subsets of referees around t

Re: [R] A combinatorial assignment problem

2014-05-01 Thread Bert Gunter
This is not really a combinatorial problem, I'll use small letters instead of caps. Arrange the r referees in a circle. start <- 1 Replicate k times{ end <- (start + m-1)%% r output: c(start,end) start <- (end+1)%% r } Bert Gunter Genentech Nonclinical Biostatistics (650) 467-7374 "Data is not

Re: [R] A combinatorial assignment problem

2014-05-01 Thread Daniel Nordlund
> -Original Message- > From: r-help-boun...@r-project.org [mailto:r-help-boun...@r-project.org] > On Behalf Of Daniel Nordlund > Sent: Thursday, May 01, 2014 1:10 AM > To: r-help@r-project.org > Subject: Re: [R] A combinatorial assignment problem > >

Re: [R] A combinatorial assignment problem

2014-05-01 Thread Daniel Nordlund
> -Original Message- > From: r-help-boun...@r-project.org [mailto:r-help-boun...@r-project.org] > On Behalf Of Ravi Varadhan > Sent: Wednesday, April 30, 2014 10:49 AM > To: r-help@r-project.org > Subject: [R] A combinatorial assignment problem > > Hi, > > I have this problem: K candidat