search

drizopoulos / JM

Joint Models for Longitudinal & Survival Data under Maximum Likelihood
34 stars 7 forks source link

Error in in optim(thetas, LogLik.splineGH, Score.splineGH, method = "BFGS", #41

Closed AAAAlicebird closed 1 year ago

AAAAlicebird commented 2 years ago

Hi, I am a beginner of Joint Models and JM package. While running jointModel() to fit joint models with competing risks, I got the following error:

Error in in optim(thetas, LogLik.splineGH, Score.splineGH, method = "BFGS", optim replies to the infinite value

My code as follow： fitLME.null <- lme(score_L ~ followyear_L, random = ~1 | id , data = data2, method = "REML") coxCRFit.pbc<- coxph(Surv(totalfollow, status2) ~(age +BMI_L+hyp_L+exersice_L)*strata+strata(strata),data = data2.idCR, x = TRUE)

jmCRFit.pbc <- jointModel(fitLME.null, coxCRFit.pbc, timeVar = "followyear_L", method = "spline-PH-aGH",
interFact = list(value=~strata, data= data2.idCR), CompRisk = T)

score_L is a scale score that ranges from 0 to 30. Status2 was the event including alive, disease and death. All subjects had at least two follow-ups with id as personal identifier.

Any information and advice you could give is much appreciated.

drizopoulos commented 1 year ago

This may have to do with the complexity of the model. You could also try the newer JMbayes2 package.