msc: Full, direct and indirect network results for MSC

In srhaile/mscpredmodel: Multiple Score Comparison of Prediction Models with Network Meta-Analysis

Description

Compute all pairwise comparisons for the full network, all direct comparisons, or all indirect.

Usage

msc_full(ps, mods = NULL, mtype = c("consistency", "inconsistency")[1],
  ref = NULL, verbose = FALSE, ...)

msc_model(ps, mods = NULL, mtype = c("consistency",
  "inconsistency")[1], ref = NULL, ...)

msc_network(ps, mods = NULL, mtype = c("consistency",
  "inconsistency")[1], ref = NULL, ...)

msc_direct(ps, mods = NULL, mtype = c("consistency",
  "inconsistency")[1], ref = NULL, verbose = FALSE, ...)

msc_indirect(ps, mods = NULL, mtype = c("consistency",
  "inconsistency")[1], ref = NULL, verbose = FALSE, ...)

## S3 method for class 'msc'
print(x, ...)

## S3 method for class 'msc'
plot(x, compare_to = NULL, newlabels = NULL, ...)

Arguments

ps	A set of raw performance estimates, see compute_performance. Since we have to compute a range of different aggregated performance measures, we start here with the raw performance estimates.
mods	A vector of variable names that are moderators, that is, covariates which could affect the differences in score performance. See also aggregate_performance. The main model in an analysis should probably not include any moderators, but they may be interesting when examining transitivity.
mtype	Type of model (default "consistency", else "inconsistency"). It is sufficient to write "c" or "i".
ref	Reference score, that is, the base score to which performance should be compared. In msc_model, the default reference score is the first one. In contrast, msc_direct and msc_indirect will print all pairwise comparisons unless a reference is specified.
verbose	If TRUE, results of each model will be printed (default FALSE)
...	In the msc functions, any other arguments are passed to rma.mv. See also consistency. In the print function, other options are passed to print.data.frame. Ignored in the plot function.
x	An object of class msc, from any of msc_{full, direct, indirect, network}.
compare_to	Deprecated. Use the ref option in the msc function instead. (If specified, only comparisons to compare_to are plotted.)
newlabels	A new vector of labels (character) for the scores can be used instead of the current vector of score names.

Value

A list of class msc, with the following components:

table

A tibble containing the model results for each pair

models

A list with all the individual pairwise models results (rma.mv)

type

"Full", "indirect" or "direct"?

performance

Label of performance score, as given in, for example, compute_performance.

Functions

• msc_full: Compute all pairwise comparisons in full network of evidence, as well as direct and indirect comparisons

• msc_model: Compute an MSC model

• msc_network: Compute all pairwise comparisons in the full network of evidence

• msc_direct: Compute all pairwise indirect comparisons

• msc_indirect: Compute all pairwise indirect comparisons

Examples

dat <- msc_sample_data()
bssamp <- get_bs_samples(dat, id, study, outcome, n.samples = 10, 
                  scores = c("a", "b", "c", "d", "e", "f"), 
                  moderators = c("age", "female", "x1"))
perf <- compute_performance(bssamp, fn = calibration_slope, lbl = "CS")
msc_model(perf, mods = "age", mtype = "inconsistency", ref = "b")
plot(nw <- msc_network(perf, mods = NULL, mtype = "inconsistency", ref = "b"))
## Not run: 
msc_direct(perf, mods = "age", mtype = "inconsistency")
msc_indirect(perf, mtype = "inconsistency")
msc_direct(perf, mtype = "inconsistency")
full <- msc_full(perf, mtype = "inconsistency")
plot(full)

## End(Not run)

srhaile/mscpredmodel documentation built on Sept. 13, 2019, 3:44 p.m.