Issue with getParDM Function When Including ID in Design Matrix for length Data Stream

[xjtusjtu]

Issue Description:
Hello,

I hope everyone's doing well. I encountered a problem while using the getParDM function and would appreciate some guidance on resolving it.

Problem Description:
When attempting to incorporate the ID (individual level effects) into the design matrix of the length data stream which follows a gamma distribution, an error message consistently pops up indicating that the parameters must have identical bounds. I'm uncertain about the appropriate steps to rectify this situation.

Context:
It's worth mentioning that I've initialized zero mass parameters for the length parameter. I suspect that this specific configuration might be contributing to the reported error.

Request:
Could someone please provide insights or suggestions on how to address this issue? Any help would be greatly appreciated.

Thank you!

Additional Information:

## fit model with 2 mixtures on TPM - transition probability matrix
Par0_mix2 <- getPar0(m1, formula = ~ voteup + ad + score + cumsum_voteup + cumsum_ad + cumsum_review + mScore, formulaPi = ~topic + price + responseRateA + responseTimeA +0, mixtures=2) 
Par0_mix2$Par$answer <- c(0.00368, 28.829, 0.001224, 0.523) # natural scale mean_1, mean_2, ...
Par0_mix2$Par$review <- c(10.276, 19.18, 0.303, 0.3864)
Par0_mix2$Par$length <- c(0.105, 8.543, 23130, 9.291, 0.99,1.e-100)
Par0_mix2$beta$beta[1:8,] <- matrix(c(6.76, 10.44, 111.06, 0.12, -0.17, 0.24, 0.04, -1.99,
                                        3.78, -5.47, -80.41, 0.10, 0.49, -0.25, 0.01, -1.21), 
                                    ncol = 2, byrow = F)
Par0_mix2$beta$beta[9:16,] <- matrix(c(6.76, 10.44, 111.06, 0.12, -0.17, 0.24, 0.04, -1.99,
                                        3.78, -5.47, -80.41, 0.10, 0.49, -0.25, 0.01, -1.21), 
                                    ncol = 2, byrow = F)
#beta0 13rows 
Par0_mix2$beta$pi <- matrix(c(2,-2,-2,2,2,-2,2,-2,2,2,
                              -3.5,3.0,-1.1),13,1,
                          dimnames=list(c("law","health","educ","finance",'tech',
                                          'mom','emotion','sports','psy','job',
                                          'price','responseRateA', 'responseTimeA'),"mix2"))

# design matrix
answerDM <- list(mu= ~ cumsum_voteup + mScore, size=~ cumsum_voteup)
reviewDM <- list(mu= ~ cumsum_voteup + mScore, size=~ cumsum_review)
lengthDM <- list(mean=~ID, sd=~ 1, zeromass=~1)

answerUserBounds <- matrix(c(0,Inf,
                             0,Inf,
                             0,Inf,
                             0,Inf
  ),nrow=2*2,byrow=TRUE,dimnames=list(NULL,c("lower","upper")))

reviewUserBounds <- matrix(c(0,Inf,
                             0,Inf,
                             0,Inf,
                             0,Inf
  ),nrow=2*2,byrow=TRUE,dimnames=list(NULL,c("lower","upper")))
lengthUserBounds <- matrix(c(0,Inf,
                             0,Inf,
                             0,Inf,
                             0,Inf,
                             0,1,0,1
  ),nrow=3*2,byrow=TRUE,dimnames=list(NULL,c("lower","upper")))
lengthUserBounds <- matrix(c(0,Inf,
                             0,Inf,
                             0,Inf,
                             0,Inf,
                             0,1,0,1
  ),nrow=3*2,byrow=TRUE,dimnames=list(NULL,c("lower","upper")))


wPar0_mix2 <- getParDM(data = Data, nbStates=2,dist=list(answer="negbinom", review="negbinom", length="gamma"),
                      Par= Par0_mix2$Par,
                      DM=list(answer=answerDM, review=reviewDM, length=lengthDM)) 

Error: Error in getParDM.default(data = Data, nbStates = 2, dist = dist, :
sorry, getParDM requires the parameters for length to have identical bounds when covariates are included in the design matrix

[xjtusjtu]

Hello everyone,

I wanted to let you know that I've successfully achieved my goal by directly using the gePar0() function, instead of the approach shown in the code above which involves using both gePar0() and geParDM() functions. If there's a better method out there, please feel free to share it with me. Thank you!