Print method does not work with spatially implicit edges

[JosiahParry]

Describe the bug

Cannot activate edges when missing geometry. This issue is based on Geographic Data Science with Python

Reproducible example

library(sf)
library(dplyr)
library(igraph)
library(sfnetworks)

gurl <- "https://raw.githubusercontent.com/gdsbook/book/master/data/cache/yoyogi_park_graph.graphml"

g <- igraph::read_graph(gurl, format = "graphml")

# covert to tbl_graph
nodes <- tidygraph::as_tbl_graph(g) 

# create notes geometry
n_geo <- as_tibble(nodes) |>
  transmute(across(c(x, y), as.numeric)) |>
  st_as_sf(coords = c("x", "y"), crs = 4326) |>
  st_geometry()

# move geometry into sfn
g_nodes <- nodes |>
  mutate(geometry = n_geo) 

g_sf <- sfnetworks::as_sfnetwork(g_nodes)
#> Checking if spatial network structure is valid...
#> Spatial network structure is valid

g_sf |>
  activate(edges)
#> # A sfnetwork with 106 nodes and 287 edges
#> #
#> Error: This call requires spatially explicit edges when applied to the edges table. Activate nodes first?

Expected behavior

I expect to have the edges be active.

R Session Info

R version 4.3.0 (2023-04-21)
Platform: aarch64-apple-darwin20 (64-bit)
Running under: macOS Ventura 13.3

Matrix products: default
BLAS:   /System/Library/Frameworks/Accelerate.framework/Versions/A/Frameworks/vecLib.framework/Versions/A/libBLAS.dylib 
LAPACK: /Library/Frameworks/R.framework/Versions/4.3-arm64/Resources/lib/libRlapack.dylib;  LAPACK version 3.11.0

locale:
[1] en_US.UTF-8/en_US.UTF-8/en_US.UTF-8/C/en_US.UTF-8/en_US.UTF-8

time zone: Europe/Berlin
tzcode source: internal

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

other attached packages:
[1] dplyr_1.1.2      sf_1.0-12        igraph_1.4.2     sfnetworks_0.6.3

loaded via a namespace (and not attached):
 [1] styler_1.9.1         utf8_1.2.3           generics_0.1.3       tidyr_1.3.0         
 [5] class_7.3-21         xml2_1.3.4           lwgeom_0.2-11        KernSmooth_2.23-20  
 [9] lattice_0.21-8       digest_0.6.31        magrittr_2.0.3       spatstat.utils_3.0-3
[13] evaluate_0.20        grid_4.3.0           bookdown_0.34        fastmap_1.1.1       
[17] R.oo_1.25.0          R.cache_0.16.0       Matrix_1.5-4         processx_3.8.1      
[21] R.utils_2.12.2       e1071_1.7-13         codegrip_0.0.0.9000  DBI_1.1.3           
[25] ps_1.7.5             purrr_1.0.1          fansi_1.0.4          scales_1.2.1        
[29] cli_3.6.1            rlang_1.1.1          crayon_1.5.2         units_0.8-2         
[33] R.methodsS3_1.8.2    reprex_2.0.2         tidygraph_1.2.3      munsell_0.5.0       
[37] withr_2.5.0          yaml_2.3.7           tools_4.3.0          sfheaders_0.4.2     
[41] colorspace_2.1-0     ggplot2_3.4.2        vctrs_0.6.3          R6_2.5.1            
[45] proxy_0.4-27         lifecycle_1.0.3      classInt_0.4-9       fs_1.6.2            
[49] callr_3.7.3          clipr_0.8.0          pkgconfig_2.0.3      pillar_1.9.0        
[53] gtable_0.3.3         glue_1.6.2           Rcpp_1.0.11          xfun_0.39           
[57] tibble_3.2.1         tidyselect_1.2.0     rstudioapi_0.14      knitr_1.42          
[61] htmltools_0.5.5      rmarkdown_2.21       compiler_4.3.0      

[idshklein]

Did you try Net |> convert(to_spatailly_explicit) |> activate (edges)?

[JosiahParry]

No idea what magic incantation that is but

g_sf |>
  tidygraph::convert(to_spatial_explicit) |> 
  activate(edges) 

works! ! !

[loreabad6]

Even if this works with that workaround, activating the edges with spatially implicit objects should still work. Will look into it, but this is probably a bug.

[loreabad6]

@JosiahParry there are a couple of issues that opened up with your example that I'd find interesting to look into. Moving this into discussions since it is not directly related to this bug.

[loreabad6]

Inside print.sfnetwork we use st_crs() to show on the header. But if no geometry column is present, st_crs cannot be called. Working on a quick fix that would activate the nodes to extract the CRS everytime, since CRS should be the same for nodes and edges. @luukvdmeer, should st_crs() work also on sfnetworks with active edges but no geometry?