bnlearn (version 3.1)

strength.plot: Arc strength plot

Description

Plot a Bayesian network and format its arcs according to the strength of the dependencies they represent. Requires the Rgraphviz package.

Usage

strength.plot(x, strength, threshold, cutpoints, highlight = NULL,
  layout = "dot", shape = "circle", main = NULL, sub = NULL,
  debug = FALSE)

Arguments

x

an object of class bn.

strength

an object of class bn.strength computed from the object of class bn corresponding to the x parameter.

threshold

a numeric value. See below.

cutpoints

an array of numeric values. See below.

highlight

a list, see graphviz.plot for details.

layout

a character string, the layout parameter to be passed to Rgraphviz. Possible values are dots, neato, twopi, circo and fdp. See Rgraphviz documentation for d

shape

a character string, the shape of the nodes. Can be either circle or ellipse.

main

a character string, the main title of the graph. It's plotted at the top of the graph.

sub

a character string, a subtitle which is plotted at the bottom of the graph.

debug

a boolean value. If TRUE a lot of debugging output is printed; otherwise the function is completely silent.

Value

• The object of class graphAM used to format and render the plot. It can be further modified using the commands present in the graph and Rgraphviz packages.

Details

The threshold parameter is used to determine which arcs are supported strongly enough by the data to be deemed significant:

• if arc strengths have been computed using conditional independence tests, any strength coefficient (which is the p-value of the test) lesser or equal than the threshold is considered significant. In this case the default value ofthresholdis equal to the value of thealphaparameter used in the call toarc.strength, which in turn defaults to the one used by the learning algorithm (if any) or to0.05.
• if arc strengths have been computed using network scores, any strength coefficient (which is the increase/decrease of the network score caused by the removal of the arc) lesser than the threshold is considered significant. In this case the default value ofthresholdis0.
• if arc strengths have been computed using bootstrap, any strength coefficient (which is the relative frequency of the arc in the networks learned from the bootstrap replicates) greater or equal than the threshold is considered significant. In this case the default value ofthresholdis0.5.

Non-significant arcs are plotted as dashed lines.

The cutpoints parameter is an array of numeric values used to divide the range of the strength coefficients into intervals. The interval each strength coefficient falls into determines the line width of the corresponding arc in the plot. The default intervals are delimited by

unique(c(0, threshold/c(10, 5, 2, 1.5, 1), 1))

if the coefficients are computed from conditional independence tests, by

1 - unique(c(0, threshold/c(10, 5, 2, 1.5, 1), 1))

for bootstrap estimates or by the quantiles

quantile(-s[s < threshold], c(0.50, 0.75, 0.90, 0.95, 1))

of the significant coefficients if network scores are used.

Examples

# plot the network learned by gs().
res = set.arc(gs(learning.test), "A", "B")
strength = arc.strength(res, learning.test, criterion = "x2")
strength.plot(res, strength)
# add another (non-significant) arc and plot the network again.
res = set.arc(res, "A", "C")
strength = arc.strength(res, learning.test, criterion = "x2")
strength.plot(res, strength)