osmbus is an R package which exports OpenStreetMap (OSM) public transport lines to GPX files.
It works with OSM relations containing the following tags :
type=routeroute=aerialway|bus|ferry|monorail|subway|train|tram|trolleybus
You can install the package from GitHub with :
# install.packages("remotes")
remotes::install_github("py-b/osmbus")The only mandatory argument is the OSM identifier of the transport line. For instance, to export this bus line (in Nancy, France) :
write_gpx(id_rel = "3220296")The data coming from the OSM server must be clean for the export to work.
As stated in the OSM documentation, the ways in the relation should be listed beginning with the way at the initial stop position and ending with the way at the terminal stop, in the right order. The package will automatically reverse some ways to produce a continuous track.
If roundabouts (or any circular ways) are part of the track, osmbus will find its way through them, i.e. it will select the only nodes used by the transport vehicle from entrance to exit.
- transport line metadata : some retrieved from OSM (for example the version of the relation) and some calculated by the package (for example the length in kilometers) ;
- coordinates and names of the stops (
role=stopin the relation), aswpt(s)elements ; - list of the track points in the correct order, as a
trksegelement.
It will be structured like this :
<?xml version="1.0" encoding="UTF-8"?>
<gpx>
<metadata>
<osm_relation id="3220296" version="79" timestamp="2020-11-08T17:23:43Z">
<colour>#20B14A</colour>
<from>Seichamps Haie Cerlin</from>
<name>Tempo 3 : Seichamps Haie Cerlin → Villers Campus Sciences</name>
<network>Stan</network>
<operator>Transdev</operator>
<ref>Tempo 3</ref>
<route>bus</route>
<to>Villers Campus Sciences</to>
<type>route</type>
</osm_relation>
<bounds minlat="48.6634437" minlon="6.1424778" maxlat="48.7196149" maxlon="6.2695481"/>
<wpt_count>46</wpt_count>
<trkpt_count>946</trkpt_count>
<trk_km>17.383</trk_km>
<copyright attribution="http://www.openstreetmap.org/copyright"/>
</metadata>
<wpt lat="48.7195418" lon="6.2657901" osm_node="3163674258" version="1">
<name>Seichamps Haie Cerlin</name>
</wpt>
<wpt lat="48.7169089" lon="6.2656034" osm_node="3163519911" version="1">
<name>Seichamps Église</name>
</wpt>
<!-- .... other way points ... -->
<wpt lat="48.6635161" lon="6.1561497" osm_node="1650185781" version="6">
<name>Villers Campus Sciences</name>
</wpt>
<trk>
<trkseg>
<trkpt lat="48.7195418" lon="6.2657901" osm_node="3163674258" version="1"/>
<trkpt lat="48.7195728" lon="6.2657585" osm_node="3163674481" version="2"/>
<trkpt lat="48.71959" lon="6.2657147" osm_node="3070509280" version="3"/>
<!-- ... other track points ... -->
<trkpt lat="48.663529" lon="6.1559462" osm_node="4455689648" version="1"/>
<trkpt lat="48.6635161" lon="6.1561497" osm_node="1650185781" version="6"/>
</trkseg>
</trk>
</gpx>Attributes of OSM objects (osm_node and version) can be ommited in
the file by setting osm_info = FALSE.
With the gpx file, you are a few steps away from creating a map like this.
(this map was not made with R but with an interactive tool called uMap)
It is also possible to get the intermediate data that was used to generate the gpx file (this can be useful to do something else than exporting to GPX).
extract_data(id_rel = "3220296")
#> $bounds
#> minlat minlon maxlat maxlon
#> 48.663444 6.142478 48.719615 6.269548
#>
#> $rel_tags
#> colour
#> "#20B14A"
#> from
#> "Seichamps Haie Cerlin"
#> name
#> "Tempo 3 : Seichamps Haie Cerlin → Villers Campus Sciences"
#> network
#> "Stan"
#> operator
#> "Transdev"
#> ref
#> "Tempo 3"
#> route
#> "bus"
#> to
#> "Villers Campus Sciences"
#> type
#> "route"
#>
#> $rel_attr
#> id version timestamp
#> "3220296" "79" "2020-11-08T17:23:43Z"
#> changeset uid user
#> "93750625" "1075986" "py_berrard"
#>
#> $stop_count
#> [1] 46
#>
#> $trkpt_count
#> [1] 946
#>
#> $trk_km
#> [1] 17.383
#>
#> $stop_base
#> # A tibble: 46 x 5
#> id name lat lon version
#> <chr> <chr> <dbl> <dbl> <chr>
#> 1 3163674258 Seichamps Haie Cerlin 48.7 6.27 1
#> 2 3163519911 Seichamps Église 48.7 6.27 1
#> 3 966281833 Donon 48.7 6.27 2
#> 4 3163208966 Jardin Roussel 48.7 6.27 1
#> 5 3140811791 Semoir 48.7 6.26 1
#> 6 963538950 Collège Goncourt 48.7 6.26 2
#> 7 3065252842 Blés d'Or 48.7 6.26 2
#> 8 962186695 Charles de Gaulle 48.7 6.25 3
#> 9 3414202098 Renaissance 48.7 6.25 2
#> 10 3138716806 Porte Verte 48.7 6.24 2
#> # ... with 36 more rows
#>
#> $trkpt_base
#> # A tibble: 946 x 5
#> id lat lon version d_last
#> <chr> <dbl> <dbl> <chr> <dbl>
#> 1 3163674258 48.7 6.27 1 0
#> 2 3163674481 48.7 6.27 2 4.16
#> 3 3070509280 48.7 6.27 3 3.74
#> 4 3163674482 48.7 6.27 2 6.68
#> 5 3070509345 48.7 6.27 3 4.95
#> 6 65074447 48.7 6.27 3 12.7
#> 7 3163674477 48.7 6.27 2 23.3
#> 8 3163674476 48.7 6.27 1 2.94
#> 9 4720935403 48.7 6.26 1 88.5
#> 10 4720935406 48.7 6.26 1 5.41
#> # ... with 936 more rows(d_last in the trkpt_base data.frame gives the distance to the
previous point, in meters)
If you wish to both export to a file and keep the data, you can affect the result of
write_gpx()to an objet :bus_data <- write_gpx(id_rel = "3220296")
The data is retrieved from OpenStreetMap using the overpass API.
You can choose which instance of the API is used via the overpass_url
argument of the functions. By default, the main instance
"http://overpass-api.de/api/interpreter" is used. See the link above
for a list of available instances.
Note that this read-only API can be a few minutes behind the editing API.