Error: Failed to create the shared library

[Xicheng Xie]

Dear all,

I encountered the same problem when I was running a linear model on HPC cluster, and I got "Error: Failed to create the shared library". I have tried the 

foo <- nimbleFunction( run = function(x = double(1)) {return(sum(x)); returnType(double())})

cfoo <- compileNimble(foo)
cfoo(1:10)

which works good on the cluster, does that mean compiler setup is working, and so what's the problem of my code?

My code is like this:

#mcmc function
run_linear_mcmc<-function(data){
  # Model Definition
  model_code <- nimbleCode({
    beta0 ~ dnorm(0, sd = 100)
    beta1 ~ dnorm(0, sd = 100)
    sigma ~ dunif(0, 50)
    for(i in 1:N) {
      mean_wbgtmax[i] ~ dnorm(beta0 + beta1 * year[i], sd = sigma)
    }
  })
 
  # set the constant
  constants<-list(N=nrow(data))
  data_list <- list(mean_wbgtmax = data$mean_wbgtmax, year = data$year)
  inits <- list(beta0 = 0, beta1 = 0, sigma = 1)
  niter<- 10000
 
 
  # configure and run MCMC
  mcmc_output<-nimbleMCMC(code = model_code,
                          constants = constants,
                          data = data_list,
                          inits = inits,
                          monitors = c('beta0', 'beta1', 'sigma'),
                          niter = niter)
 
  # get the posterior summary
  beta1_samples <- mcmc_output[,"beta1"]
 
  mean_posterior <- mean(beta1_samples)
  ci <- quantile(beta1_samples, c(0.025, 0.975))
  prob_gt_0 <- mean(beta1_samples > 0)
 
 
  return(list(ct_id=unique(data$ct_id),mean = mean_posterior, ci = ci, prob_gt_0 = prob_gt_0))
}

Defining model
Building model
Setting data and initial values
Running calculate on model
  [Note] Any error reports that follow may simply reflect missing values in model variables.
Checking model sizes and dimensions
Checking model calculations
Compiling
  [Note] This may take a minute.
  [Note] Use 'showCompilerOutput = TRUE' to see C++ compilation details.
Error: Failed to create the shared library. Run 'printErrors()' to see the compilation errors.
Execution halted


Thank you so much for your help!!

Best ,

Xicheng

[Chris Paciorek]

Hi Xicheng,

Sorry about the delay in anyone responding. Yes, your initial compilation check suggests the compiler is set up correctly.

I'm not sure why that simple model would result in a compilation error.

Please set this option first before running the code and let us know what error message you get.

nimbleOptions('showCompilerOutput' = TRUE)

--
You received this message because you are subscribed to the Google Groups "nimble-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to nimble-users...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/nimble-users/ca5d0a89-2051-4860-baee-16997e629cf6n%40googlegroups.com.