sc_segment() gives the SC index not SC0

[mdsumner]

sc_edge() is ok

silicate::sc_segment(silicate::SC0(tr))
# A tibble: 10,853 × 7
   object_ edge_  native_ .vx0   .vx1   path_ segment_
   <chr>   <chr>  <lgl>   <chr>  <chr>  <int> <chr>   
 1 erY36j  i85ZFS TRUE    aXrqMr Wc13d3     1 jrw9Mu  
 2 erY36j  i5gsKD TRUE    Wc13d3 Pp2x9a     1 8PNh6B  
 3 erY36j  VhQr0y TRUE    Pp2x9a DS3z6B     1 UlxXwD  
 4 erY36j  zi94aU TRUE    DS3z6B 2yMcaK     1 2NjKES  
 5 erY36j  RukRfJ TRUE    2yMcaK rfRHuq     1 swaWpG  
 6 erY36j  8Si0Qg TRUE    rfRHuq 1mLf2H     1 DDiH7Y  
 7 erY36j  5GVcOS TRUE    1mLf2H DQM8fd     1 lH38H9  
 8 erY36j  eWvDwS TRUE    DQM8fd kXNVWt     1 TWdXwh  
 9 erY36j  JNQCIh TRUE    kXNVWt TGRnNE     1 5LRqX2  
10 erY36j  Y0XAjf TRUE    TGRnNE MkfeJN     1 MF1Bln  
# … with 10,843 more rows
> silicate::sc_edge(silicate::SC0(tr))
# A tibble: 10,853 × 3
    .vx0  .vx1 path_
   <int> <int> <int>
 1  1821  2678     1
 2  2678  4780     1
 3  4780  5853     1

[mdsumner]

motivating use case is time intervals on trip objects, using segments()

tr <- trip(d %>% dplyr::select(X, Y, Date, Nest))
segs <- silicate::SC0(tr)
## splat out to a df for segments
segs <- bind_rows(segs$object$topology_, .id = "object") %>%
  inner_join(segs$vertex %>% transmute(vertex_ = row_number(), x0 = x_, y0 = y_, starttime = t_), c(".vx0" = "vertex_")) %>%
  inner_join(segs$vertex %>% transmute(vertex_ = row_number(), x1 = x_, y1 = y_, endtime = t_), c(".vx1" = "vertex_")) %>%
  dplyr::select(object, x0, y0, starttime, x1, y1, endtime)
segs$color_ <- palr::d_pal(segs$object)

...

 segs2 <- dplyr::filter(segs, endtime <= (dts[i]) & starttime >= (dts[i] - 3 * 24 * 3600))
  with(segs2, segments(x0, y0, x1, y1, col = color_))