bnlearn (version 3.1)

model string utilities: Build a model string from a Bayesian network and vice versa

Description

Build a model string from a Bayesian network and vice versa.

Usage

modelstring(x)
modelstring(x, debug = FALSE) <- value
model2network(string, ordering = NULL, debug = FALSE)
## S3 method for class 'bn':
as.character(x, ...)
## S3 method for class 'character':
as.bn(x)

Arguments

x

an object of class bn. modelstring (but not its replacement form) accepts also objects of class bn.fit.

string

a character string describing the Bayesian network.

ordering

the labels of all the nodes in the graph; their order is the node ordering used in the construction of the bn object. If NULL the nodes are sorted alphabetically.

value

a character string, the same as the string.

debug

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

...

extra arguments from the generic method (currently ignored).

Value

• model2network and as.bn return an object of class bn; modelstring and as.character.bn return a character string.

Details

The strings returned by modelstring have the same format as the ones returned by the modelstring function in package deal; network structures may be easily exported to and imported from that package (via the model2network function).

Examples

data(learning.test)
res = set.arc(gs(learning.test), "A", "B")
res
#
#   Bayesian network learned via Constraint-based methods
#
#   model:
#     [A][C][F][B|A][D|A:C][E|B:F]
#   nodes:                                 6
#   arcs:                                  5
#     undirected arcs:                     0
#     directed arcs:                       5
#   average markov blanket size:           2.33
#   average neighbourhood size:            1.67
#   average branching factor:              0.83
#
#   learning algorithm:                    Grow-Shrink
#   conditional independence test:         Mutual Information (discrete)
#   alpha threshold:                       0.05
#   tests used in the learning procedure:  43
#
modelstring(res)
# [1] "[A][C][F][B|A][D|A:C][E|B:F]"
res2 = model2network(modelstring(res))
res2
#
#   Random/Generated Bayesian network
#
#   model:
#     [A][C][F][B|A][D|A:C][E|B:F]
#   nodes:                                 6
#   arcs:                                  5
#     undirected arcs:                     0
#     directed arcs:                       5
#   average markov blanket size:           2.33
#   average neighbourhood size:            1.67
#   average branching factor:              0.83
#
#   generation algorithm:                  Empty
#
all.equal(res, res2)
# [1] TRUE