Skip to content

Random Effect Tobit

6 messages · Hedyeh Ahmadi, Dimitris Rizopoulos, Ben Bolker

Original
Hedyeh Ahmadi
# 
Hello all,
I was wonder if anyone is aware of an R package with random effect Tobit. Any help would be greatly appreciated.

Best,

Hedyeh Ahmadi, Ph.D.
Statistician
Keck School of Medicine
Department of Preventive Medicine
University of Southern California

Postdoctoral Scholar
Institute for Interdisciplinary Salivary Bioscience Research (IISBR)
University of California, Irvine

LinkedIn
www.linkedin.com/in/hedyeh-ahmadi<http://www.linkedin.com/in/hedyeh-ahmadi>
Dimitris Rizopoulos
# 
You can use the GLMMadaptive package (https://drizopoulos.github.io/GLMMadaptive/) and the censored.normal() family object. An example is given here: https://drizopoulos.github.io/JMbayes2/articles/Non_Gaussian_Mixed_Models.html#censored-linear-mixed-models-1 


Best,
Dimitris

-----Original Message-----
From: R-sig-mixed-models <r-sig-mixed-models-bounces at r-project.org> On Behalf Of Hedyeh Ahmadi
Sent: Tuesday, September 14, 2021 12:25 AM
To: r-sig-mixed-models at r-project.org
Subject: [R-sig-ME] Random Effect Tobit

Hello all,
I was wonder if anyone is aware of an R package with random effect Tobit. Any help would be greatly appreciated.

Best,

Hedyeh Ahmadi, Ph.D.
Statistician
Keck School of Medicine
Department of Preventive Medicine
University of Southern California

Postdoctoral Scholar
Institute for Interdisciplinary Salivary Bioscience Research (IISBR) University of California, Irvine

LinkedIn
https://eur01.safelinks.protection.outlook.com/?url=http%3A%2F%2Fwww.linkedin.com%2Fin%2Fhedyeh-ahmadi&amp;data=04%7C01%7Cd.rizopoulos%40erasmusmc.nl%7C7364e0f85ee144bb719708d97705625c%7C526638ba6af34b0fa532a1a511f4ac80%7C0%7C1%7C637671687284917965%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C2000&amp;sdata=vLU5PQqaPxCr3p6PdnRWyYdyCJhys%2BA4r9RMaS%2FnPLU%3D&amp;reserved=0<https://eur01.safelinks.protection.outlook.com/?url=http%3A%2F%2Fwww.linkedin.com%2Fin%2Fhedyeh-ahmadi&amp;data=04%7C01%7Cd.rizopoulos%40erasmusmc.nl%7C7364e0f85ee144bb719708d97705625c%7C526638ba6af34b0fa532a1a511f4ac80%7C0%7C1%7C637671687284917965%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C2000&amp;sdata=vLU5PQqaPxCr3p6PdnRWyYdyCJhys%2BA4r9RMaS%2FnPLU%3D&amp;reserved=0>



_______________________________________________
R-sig-mixed-models at r-project.org mailing list
https://eur01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fstat.ethz.ch%2Fmailman%2Flistinfo%2Fr-sig-mixed-models&amp;data=04%7C01%7Cd.rizopoulos%40erasmusmc.nl%7C7364e0f85ee144bb719708d97705625c%7C526638ba6af34b0fa532a1a511f4ac80%7C0%7C1%7C637671687284917965%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C2000&amp;sdata=xeoG32%2FqVvSSAfwGEZvNK9dHVdkdOKhfUK6ZnOH6wqo%3D&amp;reserved=0
Ben Bolker
# 
This CrossValidated answer: 
https://stats.stackexchange.com/questions/544511/zero-inflated-gaussian-for-weights-below-zero-recorded-as-0?noredirect=1#comment999636_544511

  says that the censReg package can handle random effects Tobit models 
via Gauss-Hermite Quadrature (GLMMadaptive is great, but it's nice to 
have options)
On 9/14/21 3:32 AM, Dimitris Rizopoulos wrote:

  
    

      
      
    

  


  


      

    

    

      
      

        


  

        

      Dimitris Rizopoulos
    

    

      
      #
    

  


  

      
Indeed, but the censReg does not seem to update the abscissas and weights during the optimization procedure.

Best,
Dimitris

??
Dimitris Rizopoulos
Professor of Biostatistics
Erasmus University Medical Center
The Netherlands

        

      
      
    

  
  


  


      

    

    

      
      

        


  

        

      Ben Bolker
    

    

      
      #
    

  


  

      
Can you explain a bit more? Is this the distinction that is 
sometimes made between Gauss-Hermite quadrature and *adaptive* GHQ, i.e. 
whether the expansion is centered at the population-level value or at 
zero?  (If so, I would *definitely* recommend GLMMadaptive, non-adaptive 
GHQ can be terrible in some cases ...)

   cheers
    Ben Bolker

        
On 9/14/21 9:26 PM, Dimitris Rizopoulos wrote:

            

      
      
      
    

        

      
      
    

            

      
      
      
    

        

      
      
    

            

      
      
      
    

        

      
      
    

        

  
    

      
      
    

  


  


      

    

    

      
      

        


  

        

      Dimitris Rizopoulos
    

    

      
      #
    

  


  

      
Yes, it seems to me that the non-adaptive GH is used according to procedure described in Section 3.4 here: https://cran.r-project.org/web/packages/censReg/vignettes/censReg.pdf

The adaptive GH updates the position of the abscissas and the weights during the optimization to achieve a better approximation of the log-likelihood.



??
Dimitris Rizopoulos
Professor of Biostatistics
Erasmus University Medical Center
The Netherlands