Simple feature linestring collection not visible in R leaflet

I am trying to display a simple feature collection of (disconnected) linestrings in R using sf and leaflet. The data is obtained from openstreetmap using osmdata.

Using the base plot()-function, tmap, or mapview, I am able to directly display the linestrings on a static and interactive map. However, this does not work using leaflet. Leaflet correctly centers the map, but does not display the linestrings. Fiddling with color, weight, etc. parameters in addPolyLines() does not seem to help.

Reprex

library(osmdata)
library(leaflet)
library(sf)
library(tmap)

osm <- opq(bbox = c(-0.27, 51.47, -0.20, 51.50)) %>%
  add_osm_feature(key = 'name', value = "Thames", value_exact = FALSE) %>%
  osmdata_sf()

osm$osm_lines

Simple feature collection with 2085 features and 178 fields
geometry type:  LINESTRING
dimension:      XY
bbox:           xmin: -2.029917 ymin: 51.37841 xmax: 0.6778926 ymax: 51.79032
epsg (SRID):    4326
proj4string:    +proj=longlat +datum=WGS84 +no_defs
First 10 features:
...

plot(osm$osm_lines, max.plot = 1) # OK
qtm(osm$osm_lines) + tm_lines() # OK
mapview(osm$osm_lines) # OK

leaflet(osm$osm_lines) %>% # not OK (grey)
  addPolylines()

Session-info

> library(sf)
Linking to GEOS 3.6.1, GDAL 2.2.3, proj.4 4.9.3
Warning message:
  package ‘sf’ was built under R version 3.4.4 

> sessionInfo()
R version 3.4.3 (2017-11-30)
Platform: x86_64-w64-mingw32/x64 (64-bit)
Running under: Windows >= 8 x64 (build 9200)

Matrix products: default

locale:
  [1] LC_COLLATE=English_United States.1252  LC_CTYPE=English_United States.1252    LC_MONETARY=English_United States.1252 LC_NUMERIC=C                          
[5] LC_TIME=English_United States.1252    

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

other attached packages:
  [1] tmap_1.11-2   sf_0.6-3      leaflet_2.0.1 osmdata_0.0.7

loaded via a namespace (and not attached):
  [1] httr_1.3.1         viridisLite_0.3.0  jsonlite_1.5       splines_3.4.3      geojsonlint_0.2.0  foreach_1.4.4      R.utils_2.6.0      gtools_3.5.0      
[9] shiny_1.1.0        expm_0.999-2       sp_1.3-1           stats4_3.4.3       yaml_2.1.19        LearnBayes_2.15.1  lattice_0.20-35    digest_0.6.15     
[17] RColorBrewer_1.1-2 promises_1.0.1     rvest_0.3.2        colorspace_1.3-2   plyr_1.8.4         htmltools_0.3.6    httpuv_1.4.3       Matrix_1.2-12     
[25] R.oo_1.22.0        XML_3.98-1.11      rmapshaper_0.4.0   raster_2.6-7       gmodels_2.16.2     xtable_1.8-2       webshot_0.5.0      scales_0.5.0      
[33] gdata_2.18.0       satellite_1.0.1    later_0.7.3        gdalUtils_2.0.1.14 mapview_2.4.0      magrittr_1.5       mime_0.5           deldir_0.1-15     
[41] R.methodsS3_1.7.1  nlme_3.1-131       MASS_7.3-47        xml2_1.2.0         class_7.3-14       tools_3.4.3        geosphere_1.5-7    stringr_1.3.1     
[49] V8_1.5             munsell_0.5.0      compiler_3.4.3     e1071_1.6-8        classInt_0.2-3     units_0.6-0        grid_3.4.3         tmaptools_1.2-4   
[57] RCurl_1.95-4.10    dichromat_2.0-0    iterators_1.0.9    htmlwidgets_1.2    crosstalk_1.0.0    bitops_1.0-6       base64enc_0.1-3    boot_1.3-20       
[65] codetools_0.2-15   DBI_1.0.0          jsonvalidate_1.0.0 curl_3.2           R6_2.2.2           lubridate_1.7.4    rgdal_1.3-2        rgeos_0.3-28      
[73] spdep_0.7-7        KernSmooth_2.23-15 stringi_1.1.7      osmar_1.1-7        Rcpp_0.12.17       png_0.1-7          spData_0.2.8.3     coda_0.19-1

Solution

osmdata has named features, which leaflet cannot read. See https://github.com/ropensci/osmdata/issues/100

As a hack you can set names to NULL:

osm_lines <- osm$osm_lines
names(osm_lines$geometry) <- NULL
leaflet(osm_lines) %>%
  addPolylines()