Check whether a network is connected - whether there is a path of study evidence linking every pair of treatments in the network.
is_network_connected(network)
Arguments
network: An nma_data object, as created by the functions set_*() or combine_network().
Returns
Logical TRUE or FALSE
Details
Models will still run with disconnected networks. However, estimated relative effects between treatments across disconnected parts of the network will be entirely based on the prior distribution (typically very uncertain), as there is no information to update the prior distribution. Relative effects within each connected sub-network will be estimated as if each sub-network had been analysed separately.
Examples
## Smoking cessation# Set up network of smoking cessation datahead(smoking)smk_net <- set_agd_arm(smoking, study = studyn, trt = trtc, r = r, n = n, trt_ref ="No intervention")# Print detailssmk_net
is_network_connected(smk_net)# TRUE, network is connected## A disconnected networkdisc_net <- set_agd_arm(smoking[smoking$studyn %in% c(15,21),], study = studyn, trt = trtc, r = r, n = n)is_network_connected(disc_net)# FALSE, network is disconnecteddisc_net
plot(disc_net)