Skip to content

A question about MIX package

2 messages · Li, Jia, Spencer Graves

Original
Li, Jia
# 
Hello all,
 
When I used commands "ecm.mix and dabipf.mix" to do a simulation (sample size is small 100), I got an error : Steps of ECM, missing value where True/False needed.
 
I've checked the menu, and the option "prior" of ecm.mix said that if structural zeros appear in the table, hyperparameters for those cells should be set to NA. However, it didn't say how to do that in the command. I am wondering if someone knows how to fix this.
 
I appreciate your help,
 
Jia
6 days later
Spencer Graves
# 
Have you received a reply to the post below?  I haven't seen one.  If 
not (or even if you have), I believe you could increase the chances of a 
useful reply by reading the posting guide! 
"http://www.R-project.org/posting-guide.html".  In particular, if you 
provide a sample toy example that someone else can copy from your email 
into R and experiment with modifications in less than a minute, you are 
more likely to get a useful reply.  I don't know MIX, but it sounds 
interesting, and I might be willing to experiment if I could do so 
without a major effort.  More generally, toy examples like this make it 
much easier for someone else to understand what you need.  You may think 
your question is perfectly clear.  Unfortunately, it is not as clear to 
others with a different background.

	  Good Luck.
	  spencer graves
Li, Jia wrote:

            

      
      
      
    

        
(sample size is small 100), I got an error : Steps of ECM, missing
value where True/False needed.

            

      
      
      
    

        
if structural zeros appear in the table, hyperparameters for those
cells should be set to NA. However, it didn't say how to do that in
the command. I am wondering if someone knows how to fix this.