dic: Deviance Information Criterion (DIC)

In multinma: Bayesian Network Meta-Analysis of Individual and Aggregate Data

Deviance Information Criterion (DIC)

Description

Calculate the DIC for a model fitted using the nma() function.

Usage

dic(x, penalty = c("pD", "pV"), ...)

Arguments

x	A fitted model object, inheriting class stan_nma
penalty	The method for estimating the effective number of parameters, used to penalise model fit in the DIC. Either "pD" (the default), or "pV". For survival likelihoods only "pV" is currently available.
...	Other arguments (not used)

Value

A nma_dic object.

See Also

print.nma_dic() for printing details, plot.nma_dic() for producing plots of residual deviance contributions.

Examples

## Smoking cessation

# Run smoking FE NMA example if not already available
if (!exists("smk_fit_FE")) example("example_smk_fe", run.donttest = TRUE)


# Run smoking RE NMA example if not already available
if (!exists("smk_fit_RE")) example("example_smk_re", run.donttest = TRUE)


# Compare DIC of FE and RE models
(smk_dic_FE <- dic(smk_fit_FE))
(smk_dic_RE <- dic(smk_fit_RE))   # substantially better fit

# Plot residual deviance contributions under RE model
plot(smk_dic_RE)

# Check for inconsistency using UME model


# Run smoking UME NMA example if not already available
if (!exists("smk_fit_RE_UME")) example("example_smk_ume", run.donttest = TRUE)


# Compare DIC
smk_dic_RE
(smk_dic_RE_UME <- dic(smk_fit_RE_UME))  # no difference in fit

# Compare residual deviance contributions
plot(smk_dic_RE, smk_dic_RE_UME, show_uncertainty = FALSE)

multinma documentation built on April 4, 2025, 3:46 a.m.