Function reference • multinma

Package overview

multinma-package multinma: multinma: A Package for Network Meta-Analysis of Individual and Aggregate Data in Stan

Defining a network

Setting up a network from different data sources, creating network plots.

set_agd_arm(): Set up arm-based aggregate data

set_agd_contrast(): Set up contrast-based aggregate data

set_agd_surv(): Set up aggregate survival data

set_ipd(): Set up individual patient data

combine_network(): Combine multiple data sources into one network

multi(): Multinomial outcome data

print(<nma_data>) print(<mlnmr_data>): Print nma_data objects

plot(<nma_data>): Network plots

as.igraph(<nma_data>) as_tbl_graph(<nma_data>): Convert networks to graph objects

nma_data-class nma_data mlnmr_data mlnmr_data-class: The nma_data class

is_network_connected(): Check network connectedness

Setting up numerical integration (ML-NMR only)

Multilevel network meta-regression models require numerical integration points to be specified for the distributions of covariates in each aggregate data study in the network.

add_integration() unnest_integration(): Add numerical integration points to aggregate data

distr(): Specify a general marginal distribution

qbern() pbern() dbern(): The Bernoulli Distribution

qgamma() dgamma() pgamma(): The Gamma distribution

dgent() pgent() qgent(): Generalised Student's t distribution (with location and scale)

dlogt() plogt() qlogt(): Log Student's t distribution

qlogitnorm() dlogitnorm() plogitnorm(): The logit Normal distribution

Prior distributions

Specify and summarise prior distributions.

normal() half_normal() log_normal() cauchy() half_cauchy() student_t() half_student_t() log_student_t() exponential() flat(): Prior distributions

summary(<nma_prior>): Summary of prior distributions

nma_prior-class nma_prior: The nma_prior class

plot_prior_posterior(): Plot prior vs posterior distribution

dgent() pgent() qgent(): Generalised Student's t distribution (with location and scale)

dlogt() plogt() qlogt(): Log Student's t distribution

Model fitting

Model specification and fitting is accomplished using the nma() function.

nma(): Network meta-analysis models

print(<stan_nma>): Print stan_nma objects

summary(<stan_nma>) plot(<stan_nma>): Posterior summaries from stan_nma objects

pairs(<stan_nma>): Matrix of plots for a stan_nma object

stan_nma-class stan_nma stan_mlnmr: The stan_nma class

adapt_delta: Target average acceptance probability

RE_cor() which_RE(): Random effects structure

.default() .is_default(): Set default values

Model checking and comparison

Checking model fit and comparing models.

plot_prior_posterior(): Plot prior vs posterior distribution

plot_integration_error(): Plot numerical integration error

dic(): Deviance Information Criterion (DIC)

print(<nma_dic>): Print DIC details

plot(<nma_dic>): Plots of model fit diagnostics

nma_dic-class nma_dic: The nma_dic class

loo(<stan_nma>) waic(<stan_nma>): Model comparison using the loo package

Node-splitting

Generate and summarise node-splitting models for assessing inconsistency.

get_nodesplits() has_direct() has_indirect(): Direct and indirect evidence

nma_nodesplit-class nma_nodesplit nma_nodesplit_df nma_nodesplit_df-class: The nma_nodesplit class

print(<nma_nodesplit_df>) print(<nma_nodesplit>): Print nma_nodesplit_df objects

summary(<nma_nodesplit_df>) summary(<nma_nodesplit>) plot(<nma_nodesplit>) plot(<nma_nodesplit_df>): Summarise the results of node-splitting models

nodesplit_summary-class nodesplit_summary: The nodesplit_summary class

print(<nodesplit_summary>) as_tibble(<nodesplit_summary>) as.tibble.nodesplit_summary() as.data.frame(<nodesplit_summary>): Methods for nodesplit_summary objects

plot(<nodesplit_summary>): Plots of node-splitting models

Posterior summaries and working with fitted models

Producing and plotting relative effects, absolute predictions, marginal effects, posterior ranks and rank probabilities. Converting to MCMC arrays and matrices.