Convert a Graph Object Back Into an R Database

Reverses the data.base -> graph_object direction already covered by as_igraph (and database_relationships): extracts a graph's node and edge tables into a named list of data frames (an R database), the same list(nodes = ..., edges = ...) shape used throughout this package's data.base <-> excel tooling (see write_workbook/read_workbook).

Usage

as_database(x, ...)

# Default S3 method
as_database(x, ...)

# S3 method for class 'igraph'
as_database(x, ...)

# S3 method for class 'visNetwork'
as_database(x, ...)

# S3 method for class 'list'
as_database(x, ...)

Arguments

x

A graph object: an igraph (or tbl_graph, which extends it and is dispatched via the igraph method), a visNetwork htmlwidget (e.g. from visNetwork::visNetwork() or funviewR::plot_dependency_graph()), or a plain list(nodes = ..., edges = ...) (e.g. read back from read_workbook) to be validated and stamped as one.

...

Additional arguments passed to methods.

Value

A named list list(nodes = <tibble>, edges = <tibble>) of class cbe_database – the class database_setops and compare_df dispatch on.

Details

Every node/edge attribute present on x is preserved as its own column – nothing is dropped, renamed, or flattened (e.g. a rich FilePath-derived pipeline graph's stage/mtime/ description/artifact_role, see compute_graph, all come through intact). There's no merging step here since only one graph is involved; the "whose attributes win" question only arises once you combine two databases – see database_setops for that.

Examples

g <- igraph::make_ring(5)
as_database(g)
#> $nodes
#> # A tibble: 5 × 1
#>   name 
#>   <chr>
#> 1 N_1  
#> 2 N_2  
#> 3 N_3  
#> 4 N_4  
#> 5 N_5  
#> 
#> $edges
#> # A tibble: 5 × 2
#>   from  to   
#>   <chr> <chr>
#> 1 N_1   N_2  
#> 2 N_2   N_3  
#> 3 N_3   N_4  
#> 4 N_4   N_5  
#> 5 N_1   N_5  
#> 
#> attr(,"class")
#> [1] "cbe_database" "list"