migraph (version 0.8.13)

regression: Linear regression for network data

Description

This function extends the multiple regression quadratic assignment procedure (MRQAP) of network linear model to two mode networks. It also works with combined graph/network objects by constructing the various dependent and independent matrices for the user. Lastly, because it relies on an object that contains all this information it can offer a more informative formula-based system for specifying the model.

Usage

network_reg(formula, data, ...)
# S3 method for netlm
summary(object, reps = 1000, ...)
# S3 method for summary.netlm
print(
  x,
  digits = max(3, getOption("digits") - 3),
  signif.stars = getOption("show.signif.stars"),
  ...
)

Arguments

formula

A formula describing the relationship being tested.

data

A named list of matrices, graphs, or a tidygraph object.

...

Arguments passed on to lm().

object

an object of class "netlm", usually as a result of a call to network_reg().

reps

Integer indicating the number of draws to use for quantile estimation. (Relevant to the null hypothesis test only - the analysis itself is unaffected by this parameter.) Note that, as for all Monte Carlo procedures, convergence is slower for more extreme quantiles. By default, reps=1000.

x

an object of class "summary.netlm", usually, a result of a call to summary.netlm().

digits

the number of significant digits to use when printing.

signif.stars

logical. If TRUE, <U+2018>significance stars<U+2019> are printed for each coefficient.

Examples

# NOT RUN {
messages <- mutate_edges(ison_eies, 
  generate_random(ison_eies), attr_name = "random")
model1 <- network_reg(weight ~ random + 
  same(Discipline) + same(Citations), messages)
summary(model1, reps = 200) # increase reps for publication
# }