motifs_distribution: Compute statistical properties (expectation and variance) of...

Description Usage Arguments Value Examples

View source: R/count_motifs.R

Description

This function supports the Erdős-Rényi Model (erdos_renyi) and the the Actor’s Choice Model (actors_choice). The model can be specified using the model parameter. The Erdős-Rényi Model can be used without providing further parameters. In case of the Actor’s Choice Model a level of the given network can be specified which is only level assumed to be variable. All other levels are assumed to be fixed. Per default, level = -1, the first level carrying two nodes in the signature of the motif is selected as variable level. Set the level parameter to the value of the lvl_attr of the nodes in the desired level to specify the level manually.

Usage

1
2
3
4
5
6
7
8
9
motifs_distribution(
  net,
  motifs,
  lvl_attr = "sesType",
  model = "erdos_renyi",
  level = -1,
  omit_total_result = TRUE,
  directed = NULL
)

Arguments

net

network object

motifs

list of motif identifiers describing the motifs whose distribution shall be analysed

lvl_attr

character vector specifying the attribute name where level information is stored in net.

model

baseline model to be used. options are "erdos_renyi" and "actors_choice". Defaults to "erdos_renyi".

level

Additional parameter to set the level to vary for the actors_choice model manually. All other levels are held fixed.

omit_total_result

whether total results shall be omitted

directed

whether the graph shall be treated as a directed graph. Per default (NULL), this is determined automatically using the structure of the provided network object

Value

data frame with one column giving names of motif identifers and two column giving expectation and variances per motif. For other motifs, expectations are computed but variances are returned as NaN.

Examples

1
2
3
4
## Not run: 
motifs_distribution(ml_net, motif = list("1,2[I.C]"), directed = FALSE)

## End(Not run)

motifr documentation built on Dec. 15, 2020, 5:23 p.m.