Become an expert in R — Interactive courses, Cheat Sheets, certificates and more!
Get Started for Free

network.collapse

Convert a time range of a networkDynamic object into a static network object.


Description

This function provides an easy way to sensibly collapse the time-related information in a networkDynamic object and return a plain network object with a set of vertices, edges, and attributes that appropriately correspond to those appearing in the original networkDynamic object during the query interval.

Usage

network.collapse(dnet, onset = NULL, terminus = NULL, at = NULL, length = NULL, 
                 rule = c("any", "all","earliest","latest"), active.default = TRUE, 
                 retain.all.vertices=FALSE, rm.time.info=TRUE, ...)
dnet%k%at

Arguments

dnet

A networkDynamic object with possible vertex, edge, network, and attribute spell information.

onset

optionally, the start of the specified interval. This must be accompanied by one of terminus or length.

terminus

optionally, the end of the specified interval. This must be accompanied by one of onset or length.

length

optionally, the length of the specified interval. This must be accompanied by one of onset or terminus.

at

optionally, a single time point.

rule

a text string for defining “active” for this call: any if elements active at any time during the interval are to be used, or all if elements must be active over the entire interval. The value earliest behaves like any but specifies that when multiple attribute values are encountered, only the earliest will be returned. The value latest behaves like any but specifies that when multiple attribute values are encountered, only the latest will be returned.

active.default

logical; should elements without an activity attribute be regarded as active by default?

retain.all.vertices

logical; should the extracted network retain all vertices, ignoring the vertex activity spells of x in order to ensure that the network returned has the same size as dnet?

rm.time.info

logical; if TRUE, the net.obs.period attribute will be removed (if it exists), and the activity summary attributes activity.count and activity.duration will not be attached to edges and vertices

...

Possible additional arguments (not yet invented) to handle aggregation of attributes.

Details

First performs a network.extract on the passed networkDynamic object with the specified time range to get the appropriate set of active edges. Extracts appropriate values from any dynamic attributes and stores them as static attributes and optionally (if rm.time.info=FALSE) computes some crude summary attributes for edge and vertex spells (activity.count, activity.duration),. Then removes all activity.attribute and dynamic.attributes information and returns a plain network without the networkDynamic class.

The %k% operator (‘K’ for kollapse) is a shortcut for the 'at' version of network.collapse.

If no temporal arguments are specified, defaults to collapsing entire time range of the input network (onset=-Inf,terminus=Inf). Original network is not modified.

Arbitrary attribute values may be returned when query spells with a duration are used for continuous time networks (i.e. anything other than 'at') or query spells that don't line up with the discrete units for discrete time networks are used. If a query spell intersects with multiple attribute values, a warning will be produced and only the earliest value will be used (see get.vertex.attribute.active). To avoid ambiguity (and the warning), appropriate handling can be specified by setting rule='earliest' or rule='latest' to indicate which of the multiple values should be returned.

The duration values returned for edges do not account for any potential ‘censoring’ of observations in the original network.

Value

A new 'static' network object corresponding to the specified time range of the networkDynamic argument. If the original network contained dynamic TEA attributes (i.e. 'weight.active'), the appropriate (if possible) value is queried and stored as a non-TEA attribute ('weight').

Note

This function may be quite computationally expensive if the network contains lots of attributes. For many tasks it is possible to avoid collapsing the network by using is.active, the dynamic network.extensions, and the attribute.activity.functions.

Author(s)

Skye Bender-deMoll

See Also

See also network.extract for extracting sub-ranges of a networkDynamic, get.vertex.attribute.active for more on TEA attributes, and as.network.networkDynamic for stripping the the networkDynamic class from an object without the expense of modifying or removing the activity attributes.

Examples

# create a network with some basic activity
test<-network.initialize(5)
add.edges.active(test, tail=c(1,2,3), head=c(2,3,4),onset=0,terminus=1)
activate.edges(test,onset=3,terminus=5)
activate.edges(test,onset=-2,terminus=-1)

# collapse the whole thing
net <-network.collapse(test)
is.networkDynamic(net)
get.vertex.attribute(net,'activity.duration')
get.edge.value(net,'activity.count')
get.edge.value(net,'activity.duration')

# add a dynamic edge attribute
activate.edge.attribute(test,'weight',5,onset=3,terminus=4)
activate.edge.attribute(test,'weight',3,onset=4,terminus=5)

# collapse with an interval query
net3<-network.collapse(test,onset=3,terminus=4)
get.edge.value(net3,'weight')

# note that if we use a query that intersects mutiple 
# attribute values it will generate a warning. 
# try commented line below:

# net3<-network.collapse(test,onset=3,terminus=5)

# but should be safe from attribute issues when
# collapsing with a point query
net3<-network.collapse(test,at=3)
get.edge.value(net3,'weight')

# can use operator version for point query instead
net3<-test%k%4.5
get.edge.value(net3,'weight')

networkDynamic

Dynamic Extensions for Network Objects

v0.10.1
GPL-3
Authors
Carter T. Butts [aut], Ayn Leslie-Cook [aut], Pavel N. Krivitsky [aut], Skye Bender-deMoll [aut, cre], Zack Almquist [ctb], David R. Hunter [ctb], Li Wang [ctb], Kirk Li [ctb], Steven M. Goodreau [ctb], Jeffrey Horner [ctb], Martina Morris [ctb]
Initial release
2020-01-16

We don't support your browser anymore

Please choose more modern alternatives, such as Google Chrome or Mozilla Firefox.