How to speed time while running mixtu... PreviousNext
Mplus Discussion > Latent Variable Mixture Modeling >
Message/Author
 Ali posted on Wednesday, February 12, 2014 - 2:41 pm
Hi,
I am running A mixture IRT model with 20 replication to try fit if the model fits the number of classes from 1 to 3 well.
I installed to MplusAutomation to run, so it can run automatically. But it took a lot of time to run around 1 day.Also, if i increase start value from 500 125 to 1000 250, it will take more time.
My laptop system is 32 bit. Is there is any way to speed up?
 Bengt O. Muthen posted on Wednesday, February 12, 2014 - 3:18 pm
When you do monte carlo simulations with mixtures, note that when you don't use our recommended Starts=0 the latent classes may change order over the replications making the summary results meaningless.

You should get a 64-bit machine with at least 8 processors if you do this sort of work.
 Ali posted on Wednesday, February 12, 2014 - 8:12 pm
Thank you!
But I didn't run monte carlo simulations with mixture. I just simulated the data from other program with 20 times, since I am no still familiar with monte carlo. So, I just simply run the 20 set data to fit the number of class in a mixture Rasch model to check which model fits well.
So, does it still need a 64-bit machine with at least 8 processor?
 Bengt O. Muthen posted on Friday, February 14, 2014 - 10:44 am
Probably not.
Back to top
Add Your Message Here
Post:
Username: Posting Information:
This is a private posting area. Only registered users and moderators may post messages here.
Password:
Options: Enable HTML code in message
Automatically activate URLs in message
Action: