如何在R中抑制JAGS的"value out of range"警告

I've dealt with this exact issue before—those persistent value out of range in 'lgamma' messages come directly from JAGS itself (not R's native warning system), which is why standard R suppression tools like suppressWarnings() or try(silent=TRUE) don't work. Here are two reliable solutions:

1. Temporary Redirect Standard Error (Quick Fix to Suppress Output) #

Since JAGS writes these messages to the standard error stream (stderr), we can temporarily redirect this stream to a null device while running the model, then restore it afterward. This will silence all JAGS-related stderr output without affecting your regular R warnings or messages.

library("R2jags")

# Redirect stderr to null to silence JAGS messages
sink(file = nullfile(), type = "message")

# Run your JAGS model as usual
fit <- jags(
  data = list("X.obs", "n", "N", "nMissing", "ones1", "ones2", "missRate"),
  inits = jags.inits,
  parameters.to.save = "p",
  model.file = example.model,
  n.chains = 3,
  n.iter = 1000,
  n.burnin = 500,
  n.thin = 1,
  progress.bar = "none"
)

# Restore normal stderr output
sink(type = "message")

2. Fix the Model to Eliminate Warnings at the Source #

The lgamma error happens because your model is trying to compute the log factorial of a negative number (since logfact(k) = lgamma(k+1)). This occurs when X.missing[t] exceeds n[t] (making n[t] - X.missing[t] negative). Add a truncation constraint to X.missing[t] to ensure it never exceeds n[t], and you'll eliminate the warnings entirely:

Modify the X.missing[t] line in your JAGS model to include truncation bounds:

for (t in 1:N) {
  # Truncate X.missing to be between 0 and n[t] to avoid negative logfact arguments
  X.missing[t] ~ dpois(missRate) T(0, n[t])
}

This ensures n[t] - X.missing[t] is always non-negative, so the logfact calculation won't trigger an out-of-range error.

Why Your Previous Attempts Failed #

• R's suppressWarnings()/try(silent=TRUE) only handle warnings generated by R, not messages from external programs like JAGS.
• Adding quiet=TRUE to jags.model() silences some JAGS startup messages, but not runtime calculation errors like this lgamma issue.

内容的提问来源于stack exchange，提问作者Scransom