Error in graph_to_tree(graph, mode = direction) : Graph must be directed

[rbenel]

Hi,
I have an un-directed graph which produces an image with a single central node and many vertices. The code I used two weeks ago no longer produces that image and instead I receive the error "Error in graph_to_tree(graph, mode = direction) : Graph must be directed".

When using plot from igraph I receive the same image that I previously received using ggraph

I updated ggraph and I still nothing, any ideas?

`> sessionInfo()
R version 3.6.3 (2020-02-29)
Platform: x86_64-pc-linux-gnu (64-bit)
Running under: Ubuntu 18.04.4 LTS

Matrix products: default
BLAS: /usr/lib/x86_64-linux-gnu/blas/libblas.so.3.7.1
LAPACK: /usr/lib/x86_64-linux-gnu/lapack/liblapack.so.3.7.1

locale:
[1] LC_CTYPE=en_US.UTF-8 LC_NUMERIC=C LC_TIME=en_US.UTF-8 LC_COLLATE=en_US.UTF-8
[5] LC_MONETARY=en_US.UTF-8 LC_MESSAGES=en_US.UTF-8 LC_PAPER=en_US.UTF-8 LC_NAME=C
[9] LC_ADDRESS=C LC_TELEPHONE=C LC_MEASUREMENT=en_US.UTF-8 LC_IDENTIFICATION=C

attached base packages:
[1] stats graphics grDevices utils datasets methods base

other attached packages:
[1] viridis_0.5.1 viridisLite_0.3.0 ggrepel_0.8.2 ggraph_2.0.3 igraph_1.2.5 corrr_0.4.2
[7] forcats_0.5.0 stringr_1.4.0 purrr_0.3.4 readr_1.3.1 tibble_3.0.1 tidyverse_1.3.0
[13] ggplot2_3.3.1 data.table_1.12.8 reshape_0.8.8 pheatmap_1.0.12 tidyr_1.1.0 dplyr_1.0.0

loaded via a namespace (and not attached):
[1] Rcpp_1.0.4.6 lubridate_1.7.8 lattice_0.20-41 assertthat_0.2.1 digest_0.6.25
[6] ggforce_0.3.1 R6_2.4.1 cellranger_1.1.0 plyr_1.8.6 backports_1.1.7
[11] reprex_0.3.0 httr_1.4.1 pillar_1.4.4 rlang_0.4.6 readxl_1.3.1
[16] rstudioapi_0.11 blob_1.2.1 Matrix_1.2-18 BiocParallel_1.18.1 polyclip_1.10-0
[21] munsell_0.5.0 broom_0.5.6 compiler_3.6.3 modelr_0.1.8 pkgconfig_2.0.3
[26] tidyselect_1.1.0 gridExtra_2.3 graphlayouts_0.7.0 fansi_0.4.1 crayon_1.3.4
[31] dbplyr_1.4.4 withr_2.2.0 MASS_7.3-51.6 grid_3.6.3 nlme_3.1-147
[36] jsonlite_1.6.1 gtable_0.3.0 lifecycle_0.2.0 DBI_1.1.0 magrittr_1.5
[41] scales_1.1.1 cli_2.0.2 stringi_1.4.6 farver_2.0.3 fs_1.4.1
[46] xml2_1.3.2 ellipsis_0.3.1 generics_0.0.2 vctrs_0.3.0 RColorBrewer_1.1-2
[51] tools_3.6.3 glue_1.4.1 tweenr_1.0.1 hms_0.5.3 parallel_3.6.3
[56] yaml_2.2.1 colorspace_1.4-1 tidygraph_1.2.0 rvest_0.3.5 haven_2.3.1`

[deziahuja]

I'm encountering the same error message when I work with my undirected, egocentric networks. Has there been any work towards resolving this issue?

[rbenel]

I used igraph instead, as the error didn't seem to be fixed.

[beanumber]

FYI, I was able to get past this by specifying the layout. So instead of

ggraph(g)

use

ggraph(g, 'tree')

[thomasp85]

if anyone can provide a reprex I'd be happy to look into it

[demiliosaldana]

FYI, I was able to get past this by specifying the layout. So instead of

ggraph(g)

use

ggraph(g, 'tree')

Hi @beanumber! I´m wondering if you could tell me where did you find information about "mode" in the "ggraph" function, in order to avoid the Error. I´ve been searching about this "modes", but I couldn´t find anything related it.
Thank you Ben! Have a good one!

Emilio

[beanumber]

@demiliosaldana: mode is an argument to the internal function graph_to_tree().

ggraph:::graph_to_tree
#> function (graph, mode) 
#> {
#>     if (!is.directed(graph)) {
#>         stop("Graph must be directed")
#>     }
#>     graph <- simplify(graph, edge.attr.comb = "first")
#>     parent_dir <- if (mode == "out") 
#>         "in"
#>     else "out"
#>     comp <- components(graph, "weak")
#>     graphs <- lapply(seq_len(comp$no), function(i) {
#>         graph <- induced_subgraph(graph, which(comp$membership == 
#>             i))
#>         n_parents <- degree(graph, mode = parent_dir)
#>         if (!any(n_parents == 0)) {
#>             stop("No root in graph. Provide graph with one parentless node")
#>         }
#>         if (any(n_parents > 1)) {
#>             message("Multiple parents. Unfolding graph")
#>             root <- which(degree(graph, mode = parent_dir) == 
#>                 0)
#>             if (length(root) > 1) {
#>                 message("Multiple roots in graph. Choosing the first")
#>                 root <- root[1]
#>             }
#>             tree <- unfold_tree(graph, mode = mode, roots = root)
#>             vattr <- lapply(vertex_attr(graph), `[`, i = tree$vertex_index)
#>             vertex_attr(tree$tree) <- vattr
#>             graph <- tree$tree
#>         }
#>         as_tbl_graph(graph)
#>     })
#>     do.call(bind_graphs, graphs)
#> }
#> <bytecode: 0x563400c8e110>
#> <environment: namespace:ggraph>

^{Created on 2021-01-05 by the reprex package (v0.3.0)}

[mschubert]

if anyone can provide a reprex I'd be happy to look into it

I can reproduce this on a "tree" graph with edge fields from and to but no edges (i.e, the edge data.frame has no rows):

net = structure(list(1, FALSE, numeric(0), numeric(0), numeric(0),
    numeric(0), c(0, 0), c(0, 0), list(c(1, 0, 1), structure(list(), .Names = character(0)),
        list(name = "node"), list(weight = numeric(0)))),
        class = c("tbl_graph", "igraph"), active = "nodes")
    

ggraph(igraph::upgrade_graph(net))
# Error in graph_to_tree(graph, mode = direction) : Graph must be directed

[thomasp85]

I think this is fixed now but without a reprex for the original issue it is hard to say