impute.network.list {tergm}

Impute missing dyads in a series of networks

Description

This function takes a list of networks with missing dyads and returns a list of networks with missing dyads imputed according to a list of imputation directives.

Usage

impute.network.list(
  nwl,
  imputers = c(),
  nwl.prepend = list(),
  nwl.append = list()
)

Arguments

nwl

A list of network objects or a network.list object.

imputers

A character vector giving one or more methods to impute missing dyads. Currenly implemented methods are as follows:

next

Impute the state of the same dyad in the next network in the list (or later, if that one is also missing). This imputation method is likely to lead to an underestimation of the tie-change rates. The last network in the list cannot be imputed this way.

previous

Impute the state of the same dyad in the previous network in the list (or earlier, if that one is also missing). The first network in the list cannot be imputed this way.

majority

Impute the missing dyad with the value of the majority among the non-missing dyads in that time step's network. A network that has exactly the same number of ties as non-missing non-ties cannot be imputed this way.

0

Assume missing dyads are all non-ties.

1

Assume missing dyads are all ties.

If length(imputers)>1 the specified imputation methods will be applied in succession. For example, imputers=c("next","previous","majority","0") would first try to impute a missing dyad with the next time step's value. If it, and all of the later values for that dyad are missing, it will try to impute it with the previous time step's value. If it, and all of the earlier values for that dyad are missing as well, it will try to impute it with the value of the majority of non-missing dyads for that time step. If there is an exact tie, it will impute 0.

nwl.prepend

An optional list of networks to treat as preceding those in nwl. They will not be imputed or returned, but they can be useful for imputing dyads in the first network in nwl, when using "previous" imputer.

nwl.append

An optional list of networks to treat as following those in nwl. They will not be imputed or returned, but they can be useful for imputing dyads in the last network in nwl, when using "next" imputer.

Value

A list of networks with missing dyads imputed.

See Also

network, is.na