Simutrans TimetablesCZ · fork 122.0 · pak128 VZ
mirror/pipeline/plzen-bus/README.mdView on GitHub ↗

Plzeň buses (PMDP)

Published at https://claude.ai/artifact/LE4trKFDZjkAjpX2EhvJCa (republish to the same URL after a rebuild). Maintained by session simutrans-timetable-a2.

The Plzeň bus sheet. It has no scripts of its own: scripts/run.py runs the Plzeň trolleybus pipeline's scripts (../plzen-trolleybus/scripts) with --base pointing here, in the way the Brno bus pipeline builds on ../brno-trolleybus. The raw data (PMDP GTFS, seznam-autobusu.cz day overviews, per-line lists, line histories, PMDP register) is stored once in ../plzen-trolleybus/data/raw and read from there. The shared rules are in ../README.md; the method, sources and fetch steps are in ../plzen-trolleybus/README.md.

State file Network Vehicles Output (project root)
states/2026-09-29.json Tue 29 Sep 2026 (GTFS service day 29 Sep) Fri 18 Sep 2026 Plzeň Bus Peak Fleet.html 
cd pipeline/plzen-bus/scripts
python run.py --state 2026-09-29

This folder holds: states/ (the bus state), data/derived// (timetable, fleet), assets/pak_map.json, pak_request.md and this README. Built by session simutrans-timetable-a2 on 29 Sep 2026; not published yet.

Scope

Every PMDP city bus line with afternoon-peak service (≥ 3 trips in 15:00–18:00 and a 16:xx departure): 20, 21, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 38, 39, 40, 41, 42, 51. Left out:

  • 23, 36, 37, 43, 50, 58: no trip in 15:00–18:00; 56 (Vinice – Město Touškov): one each way;
  • night lines N1–N6, N11, N12.

Subcontractors. The PMDP GTFS has one agency, and every bus logged on 18 Sep resolved to the PMDP register (04_fleet.py would show another operator's car with its operator and the VZ-Plzen-bus package; 01_fetch.py --operator autobus fetches that operator's register). None ran a PMDP line in the peak, so there is nothing to request for them.

Decisions

  • Day and network: as for the trolleybuses: vehicles Fri 18 Sep 2026 (23–25 and 29 Sep have 0–4 bus records), network Tue 29 Sep 2026.
  • Circuits and branches (from propose.py --route-type 3):
    • 28 alternates between Papírna Bukovec and Bukovec: one line with CAN Husova twice (turn: 28|CAN Husova, since both are turning points).
    • 33 + 40 form a loop: 33 goes out by the Lochotín hospital and 40 comes back by Lochotín, and the other way round (8 of 10 and 8 of 11 in the chaining). Two Simutrans lines, one per direction: 33 (33 out, 40 back) and 40 (40 out, 33 back). 40 has no record since 31 Aug: since September its buses are logged as 33 kurzy, so both lines share the 33 kurzy (records "33").
    • 42 has one record ever (Dec 2024), while 26 is logged with 5 weekday kurzy and needs only 2. Both run Bory – Lhota. Inferred and marked "not confirmed" on the sheet: 42 runs on 26 kurzy (records "26"; 42 takes the highest numbers).
    • Branches as their own lines: 26N (Bory – Nová Ves, hourly), 30P (Borská pole – Sídliště Košutka, every 30), 41K (Vinice – Křimice, every 30). Their kurzy are the highest of the record line (a choice).
    • 24 and 25 (both Bory – Bolevec, different routes) stay separate: each has its own kurzy.
    • 32 Škoda VIII. brána → NC Černice (2 trips in 15–18 h) is not a peak trip.
  • Stop types: every entry is a turning point (−1). No through stops on these lines.
  • Hand-set rows (method check of 25 Oct 2026, pipeline/brno/method-check-2026-10-25.md; fit.overrides, key |||, notes on each line):
    • 26 one bus an hour at both ends (rate: 1; Lhota :54, the turn of the 16:30 bus): the 16:00-hour fit had given Lhota 2 an hour against 1 from Bory.
    • 25 Bolevec :00, the turn of the hourly Bory :07 trips (its 16:15 fit was a single departure).
    • 29 every 15 at both ends (rate: 4), 33 and 40 every 15 in both directions (33 rate: 4), 38 every 30 (rate: 2): their peak forms, which need the same buses; the 16:00 hour is where they thin out.
  • Need = max(cycle, on the road per record line), as on the trolleybus sheet. On 18 Sep: 69 counted, 14 not needed, none short, none designed. The designed-vehicle path (history → real cars, roster types first, leftovers "not needed") was tested on 42 with 26's history and works; it is not needed with the 26 pool.
  • Pak: no PMDP bus objects exist; seven types are "requested" in VZ-PMDP-bus, in two liveries (pmdpcervenobila for SOR NB 12, Urbino 18 III/IV and the midibuses; pmdpcervenocernobila for SOR NS 12 and NS 18). Advert wraps map to the normal livery. Articulated: -A + -B (the Brno bus convention).

How to update

  1. Fetch new data with the trolleybus pipeline (../plzen-trolleybus/scripts/run.py --state … --fetch), then the bus lists of the new vehicle day: python run.py --state 2026-09-29 --fetch here (per-line lists and registers into ../plzen-trolleybus/data/raw).
  2. Copy the state, set the dates (the same vehicle day as the trolleybus sheet), and run python ../../plzen-trolleybus/scripts/03_timetable.py --state --base .. to check rates, approximations and not modelled.
  3. Re-check the 40 and 42 pools: if seznam-autobusu.cz logs 40 or 42 again, give them their own records.
  4. python run.py --state . For a short line, fetch its history first (01_fetch.py --history --since in the trolleybus scripts).