Timetable 2017 (GTFS) - Daten | Open-Data-Plattform öV Schweiz
Hello, could you please comment on the differences between your GTFS dataset and the one from http://gtfs.geops.ch/.
In particular, in the routes.txt file, your dataset contains a bit more than 3000 routes, while geops dataset contains around 50000 routes. Why such a big difference?
Die Daten für 2017 scheinen unvollständig, z.B. fehlen viele Haltestellen in stops.txt.
@dragoon @stationsapp Danke für den Hinweis. Wir gehen dem Thema nach und melden uns baldmöglichst wieder bei Ihnen!
@stationsapp @dragoon Hallo zusammen! Danke nochmals für euren Hinweis. Es sind nun neue Daten für 2017 online! Schaut es euch doch mal an und gebt uns Feedback! Danke!
Danke für’s Update.
stops.txt sieht gut aus, zumindest gemäss meinen Stichproben.
routes.txt scheint mir aber immer noch ein bissel mager.
und agency.txt enthält nur die SBB.
Danke für das Feedback@stationsapp!
Es ist richtig, dass das agency.txt nur die SBB enthält.
Es gibt zwei Möglichkeiten, die Agency.txt zu füllen:
Unternehmer(namen) können ausgegeben werden, wenn diese einem Betriebszweig zugeordnet sind. Hier gilt jedoch “nur ein Unternehmer pro Betriebszweig”. Da es mehrere Unternehmer pro Betriebszweig gibt, würde das bei uns nicht gehen. In unserem Fall wird deshalb der Verbund/das Unternehmen in DIVA Config gepflegt, dann wird nur diese eine Agency ausgegeben.
Bezüglich dem routes.txt File sind wir dankbar, wenn Sie konkrete Angaben liefern, dann können wir dem gerne nachgehen.
The geops dataset treats each individual trip as a route. That is incorrect. The figure of 3000 in this data is more likely correct.
A route is meant to be equivalent to a ‘line’ as presented in printed timetables. For example
S3 Aarau–Lenzburg–Othmarsingen–Zürich HB is a route. It leaves Aarau at 5.16, 6.17, 7.17 etcetera. These are the trips each day. The geops dataset treats each of these trip as a route.
Thanks for sharing the data in GTFS! I noticed that the transfers.txt is missing in the zip’s. Transfers are an integral part of GTFS and needed for routing applications. They provide information about the time needed for transfers between neighoring stations and station-internal transfers. Transfers have to be converted from the HAFAS tables METABHF and UMSTEIGB.
Could you please include transfers.txt in your conversion job and update the donwloadable zip’s?
Thanks for your question @daniel-meyer1zkb-ch .
It is true, we are not providing the Transfers.txt yet. Concerning this topic we are in contact with Mentz. We try to provide you the Transfers.txt.file as soon as possible. However, the realization is not yet secured. Thanks for your understanding and patience.
Thanks for getting back to me. In the meantime I converted the transfers.txt from the HRDF tables myself. Still, I urge to include it in the feed.
Now, I’ve came across another problem. It seems to me that the weekdays in the calendar.txt are incorrectly set. I think you shifted the input dates (according to the Verkehrstagenummer in the *A VE rows in FPLAN and the corresponding entries in BITFELD) by two days.
You can check this with the following example: service_id = “TA+b01qc” and respective trips, e.g. “1.TA.21-819-j17-1.1.R” or “1.TA.21-822-j17-1.1.R”.
The service_id in the calendar.txt says that those trips are provided tuesday through sunday, but in fact they are provided monday through friday.
Here’s the respective extract of calendar.txt:
service_id,monday,tuesday,wednesday,thursday,friday,saturday,sunday,start_date,end_date
“TA+b01qc”,“0”,“0”,“1”,“1”,“1”,“1”,“1”,“20161211”,“20171209”
The two sample trips I mentioned are 1) Kartause Ittingen - Frauenfeld, Bahnhof starting at 06:58 and 2) Oberneunforn, Gemeindehaus - Frauenfeld, Bahnhof starting at 05:58. sbb.ch states correctly for both of the mentioned sample trips: runs Mo - Fr.
Could you please verify whether my findings are valid and possibly provide a rectified calendar.txt? Thanks for taking the time to study my suggestions!
Thanks for your hint. It seems that you found this issue in the feed of the 03.02. file. There was an errror indeed. Meanwhile corrections were made and in the newest feed from 04.03. the weekdays are correct.
“21-819-j17-1”,“TA+b18sl”,“6.TA.21-819-j17-1.1.R”,“Frauenfeld, Bahnhof”,“1”
“6.TA.21-819-j17-1.1.R”,“06:58:00”,“06:58:00”,“8506869”,“1”,“0”,“0”
“6.TA.21-819-j17-1.1.R”,“06:59:00”,“06:59:00”,“8573300”,“2”,“0”,“0”
“6.TA.21-819-j17-1.1.R”,“06:59:00”,“06:59:00”,“8573323”,“3”,“0”,“0”
“6.TA.21-819-j17-1.1.R”,“07:09:00”,“07:09:00”,“gen:32000:73375:0:N”,“4”,“0”,“0”
“TA+b18sl”,“1”,“1”,“1”,“1”,“1”,“0”,“0”,“20170201”,“20171206”
I recomend you to use the newest file so this problem won`t appear anymore.
Hope that helps!
Hello, are there any plans to add shapes.txt file to the data?
Then it will be possible to generate visualisations like the ones from https://github.com/cmichi/gtfs-visualizations.
Why are irregular stops filtered out of stops.txt (e.g. 140 / Gotthard-Panoramastrecke)? GEOPS have them in their dataset and they are also shown in trip results on sbb.ch.
Hi dragoon! No there are no plans yet, sorry!
Hallo
Hi @hendrik-kobergsiemens
It`s true, this data is not provided on the platform.
They are added after the export.
You can find this additional Information in the BHFART file.
Hallo. Kann es sein, dass es aktuell nicht möglich ist, unter den Zügen (route_type=2) eine S-Bahn zu erkennen und diese entsprechend als solche korrekt zu bezeichnen (z.B. “S 24”)? Oder übersehe ich etwas?
Hallo
Es scheint als wäre stop_times.txt unvollständig. Wir können für viele Trips keine Stoptimes finden.
Hallo @sinan-yasargilglue-c & @stationsapp
In den letzten Tagen waren einige GTFS Files unvollständig oder fehlerhaft. Dies sollte nun behoben sein.
Warum werden seit dem 11.7.17 invalide routentypen (gemäss gtfs reference ) wie z.B. 1300 in routes.txt ausgegeben?