Skip to contents

Compare Heterogeneous Transition Networks

Source: R/compare_htna.R
compare_htna.Rd

Computes the descriptive network comparison implemented by Nestimate::compare_model() while retaining the HTNA actor partition. Comparing two individual networks returns one htna_comparison. Passing an htna_group as x with y = NULL compares every cohort pair and returns an htna_comparison_group; callers do not need to select or subset cohorts.

Usage

compare_htna(x, y = NULL, ...)

Arguments

x

An htna network or an htna_group.

y

A second htna network. Must be NULL when x is an htna_group, in which case all cohort pairs are compared.

...

Additional arguments passed to Nestimate::compare_model(), including scaling, measures, and network.

Value

For two networks, an object inheriting from htna_comparison and net_comparison. In addition to the Nestimate comparison fields it contains:

  • node_groups and actor_levels: the aligned actor partition;

  • edge_metrics$source_actor and edge_metrics$target_actor;

  • actor_pair_metrics: difference summaries for every directed actor-type block;

  • models: the two original, unmodified HTNA networks.

For an htna_group, a named htna_comparison_group containing every pairwise comparison is returned.

Details

Networks may contain different subsets of the declared nodes. Their weight matrices are aligned to the union of nodes, in canonical actor order, and missing nodes are represented by zero rows and columns. Shared nodes must have the same actor assignment in both networks.

See also

Nestimate::compare_model(), plot_htna_diff()

Examples

data(human_ai)
grouped <- build_htna(human_ai, actor_type = "actor_type", group = "phase")
#> Warning: A network with one long sequence is not recommended and can't be validated using bootstrap and other confirmatory testings.
#> Metadata aggregated per session: ties resolved by first occurrence in 'cluster' (24 sessions), 'actor_type' (9 sessions)
#> Warning: A network with one long sequence is not recommended and can't be validated using bootstrap and other confirmatory testings.
#> Metadata aggregated per session: ties resolved by first occurrence in 'session_date' (1 sessions), 'cluster' (18 sessions), 'actor_type' (15 sessions)
comparisons <- compare_htna(grouped)
names(comparisons)
#> [1] "Late vs Early"