Page 1 of 1
Info Dump 27/12/2014 at 16:46 #67107 | |
Danny252
![]() 1461 posts |
I've been googling and fiddling around for a few days on timetables, and given how many other people are working on this, I may as well share the information! Train IDs Trains are identified by a 3 digit "Run Number", which is tied to the shift of the driver/engineer. The first digit indicates his home depot, and the second two are the order in which his shift ends, with the first shift to end having the lowest number and the last shift to end the highest (this is due to the heavily influence of the unions and the method shift allocation, which relies on drivers choosing what shifts they would prefer and seniority having a big influence). This means that run numbers are not sequential in any way - it is entirely possible to have a shift 514 that comes on duty before shift 504, but also finishes later (due to 514 performing two round trips instead of 504's one, or similar). However, the first digit is generally a very good indicator of the line, with the exception of some early morning Orange-Brown line through trains, which retain the Orange Line Midway terminal number on the Brown Line. Home depots relevant to us are as follows: [table] [tr] [td]000[/td] [td]Green[/td] [td]Harlem[/td] [/tr] [tr] [td]300[/td] [td]Pink[/td] [td]54th[/td] [/tr] [tr] [td]400[/td] [td]Brown[/td] [td]Kimball[/td] [/tr] [tr] [td]500[/td] [td]Purple[/td] [td]Howard[/td] [/tr] [tr] [td]600[/td] [td]Green[/td] [td]Ashland[/td] [/tr] [tr] [td]700[/td] [td]Orange[/td] [td]Midway[/td] [/tr] [/table] Non-Passenger workings So far I've only come across two regular non-passenger (or, at least, non-advertised) services, which are evening Brown Line services that presumably run loaded one way and ECS on the return. Timetables I have written a program to download data from the CTA TrainTracker API at five minute intervals. This lists the train's line, run number, destination, and ETA at its next expected stop. I've also pulled out the delay flag, but this doesn't seem to be used (or I'm parsing it wrongly!). Additionally, I have put together a full set of timetables for the loop from the Goroo pages, having parsed the timings automatically. From these, I have (attempted!) to work out timings to the half minute for the various schedules used (e.g. night, morning peak, day, evening peak). I've been working on combining the data from the TrainTracker API and the timetables to assign run numbers to each timetable. This has generally worked well, although I only started collecting data at 13:45 (Chicago Local Time) on Friday, so we'll need to wait until Monday to get the morning peak data! Train Lengths This is based off this source: http://forum.chicagobus.org/topic/3319-train-lengths/ [table] [tr] [td]Line[/td][td]OffPeak[/td][td]OnPeak[/td][td]Note/Peak Times[/td] [/tr] [tr] [td]Pink[/td][td]4[/td][td]4[/td] [/tr] [tr] [td]Purple[/td][td][/td][td]6[/td][td]Express Services, Peak Only M-F[/td] [/tr] [tr] [td]Green[/td][td]4[/td][td]6[/td][td]0600-0900, 1430-1745 M-F[/td] [/tr] [tr] [td]Brown[/td][td]4[/td][td]8[/td][td]0530-1000, 1430-1930 M-F[/td] [/tr] [tr] [td]Orange[/td][td]4[/td][td]8[/td][td]0530-1000, 1430-1930 M-F[/td] [/tr] [/table] Files and links TrainTracker API data (zipped) https://drive.google.com/file/d/0B3d0C_0MNAaIb1NPNGZzN3lVckE/view?usp=sharing Hopefully this is fairly easy to understand. The two locations are advertised destination (for all except Green Line trains this is "Loop" when running inbound) and next stop, and the time is the predicted arrival at the next stop. (Note that the formatting in the first four files is different as I was still playing around with my code. I'll update the .zip in a few days with what is hopefully a full day's worth of data) Timetables https://docs.google.com/spreadsheets/d/1ARAIW2FHxyvZBHzh-P50AC__D2s2vZeUE2FqAL1HWQI/edit?usp=sharing Top line: Run Number, where known. Middle lines: Timetable data from Goroo Num cars: My estimated number of cars, not very technical. Final lines: The station to station timings, and the overall time the train is in the loop. I use the overall time to choose which set of timings is being used. Timings https://docs.google.com/spreadsheets/d/17FSGu_eprnbG2ZEYZi87O42fJlk7daIb3qYDmKk3H54/edit#gid=1883852988 My estimated timings. Each line indicates the time taken from the previous station. The "Min" column is the minimum timing, taken from the overnight times usually. The "x Allow" columns show any additional time between stations that is allocated for certain times. For example, on the Brown Line an entry of "1+0H" at LaSalle in the MP Allow column indicates that the time taken from H.W. Library to LaSalle during the Morning Peak is 1 minute, including a path/performance allowance of half a minute (which would be indicated as a "0H" path allowance at H.W. Library in UK practice). Columns after the space are the raw timetable data I used to generate the timings. Run Numbers https://docs.google.com/spreadsheets/d/1Jxm8fX6VSOtq_r2E_12cx3Dzbaa_t9cYBgTINq32pFU/edit?usp=sharing Listing of Run Numbers I have allocated so far. My method for doing these is to open an API Data File, and filter out all runs that are timetabled finish more than 5 minutes before the time of the API data file. I would then look for any run number which was not present, implying that it had just started a new service, and apply this number to the next empty column on the Timetables spreadsheet for that line. For example, if looking at the data file for 16:05:00, I would filter the End Times column with ">=16:00:00" - any standard spreadsheet software should do this, but Google Docs' filter doesn't seem able to do anything but a simple text match, so you'll probably want to download the file. Notes on the data: Naturally, as the API data is from real trains, it doesn't necessarily match the timetables exactly! A couple of minutes either way seems pretty normal. There are a few missing trains (notably some Brown Line services I have marked as "512?" a couple of times), and it seems that some incident occurred at the northern end of the Purple Line Express route leading to some trains still showing as running long after they should've terminated (perhaps my timetable data is wrong, but some trains were sat showing the same station for quite a long time). Errors and Omissions Expected! Last edited: 27/12/2014 at 16:48 by Danny252 Log in to reply The following user said thank you: leigh |
Info Dump 27/12/2014 at 17:08 #67114 | |
guidomcc
![]() 246 posts |
Are you making this into a SimSig WTT? Or is this data available for someone to use? ![]() Log in to reply |
Info Dump 27/12/2014 at 17:45 #67118 | |
Danny252
![]() 1461 posts |
" said:Are you making this into a SimSig WTT? Or is this data available for someone to use? :)That was my plan, but given that I'd seen two or three other people all say "I'll work on a timetable", I thought it better to contribute what I had to other's efforts! Log in to reply The following user said thank you: guidomcc |
Info Dump 27/12/2014 at 19:59 #67127 | |
guidomcc
![]() 246 posts |
Fair enough! This will be very useful in making my current one less.. Awful. Amazing work!
Log in to reply |
Info Dump 31/12/2014 at 12:08 #67241 | |
Danny252
![]() 1461 posts |
For some curious reason, I'm unable to edit my own earlier posts! Anywho, a whole day of raw data for 2014/12/30 can be found at this link: https://drive.google.com/file/d/0B3d0C_0MNAaIWXVXRjhBUm5vZVU/view?usp=sharing Log in to reply The following user said thank you: leigh |
Info Dump 31/12/2014 at 12:30 #67242 | |
Steamer
![]() 3958 posts ![]() |
" said:For some curious reason, I'm unable to edit my own earlier posts!As far as I can tell, there's a 24 hour time limit on edits, from when the post was made/last edited. "Don't stress/ relax/ let life roll off your backs./ Except for death and paying taxes/ everything in life.../ is only for now." (Avenue Q) Log in to reply |
Info Dump 01/01/2015 at 08:55 #67261 | |
leigh
![]() 51 posts |
" said:For some curious reason, I'm unable to edit my own earlier posts!I'm having trouble opening the .dat files in the zip. What program should I be using to read these files? Log in to reply |
Info Dump 01/01/2015 at 09:08 #67262 | |
BarryM
![]() 2158 posts |
These file names as presented, are not accessible. Maybe a Google block. Barry Barry, Sydney, New South Wales, Australia Log in to reply |
Info Dump 01/01/2015 at 09:35 #67263 | |
Peter Bennet
![]() 5377 posts ![]() |
" said:" said:Auto opened in wordpad for me and could be read OK.For some curious reason, I'm unable to edit my own earlier posts!I'm having trouble opening the .dat files in the zip. What program should I be using to read these files? Peter I identify as half man half biscuit - crumbs! Log in to reply |
Info Dump 01/01/2015 at 11:26 #67266 | |
BarryM
![]() 2158 posts |
" said:These file names as presented, are not accessible. Maybe a Google block.I, we fell for the three card trick. If you cannot see a menu, move your curser up to the top of the page and all is revealed. Barry Barry, Sydney, New South Wales, Australia Log in to reply |
Info Dump 01/01/2015 at 13:13 #67269 | |
Danny252
![]() 1461 posts |
Yes, they're all just simple text files - Notepad or Wordpad will do. You might be able to sense that I'm more used to handling data in Linux than Windows!
Log in to reply |
Info Dump 01/01/2015 at 22:03 #67286 | |
leigh
![]() 51 posts |
" said:Yes, they're all just simple text files - Notepad or Wordpad will do. You might be able to sense that I'm more used to handling data in Linux than Windows!All good now. Turned out there were 2 blank files due to no trains in that time period, and they were the 2 I tried to open first. I've spent a bit of time in the last few days watching the CTA Train Tracker map so will match both my findings with your files, and see how they line up. Log in to reply |
Info Dump 01/01/2015 at 23:22 #67291 | |
Muzer
![]() 718 posts |
" said:Yes, they're all just simple text files - Notepad or Wordpad will do. You might be able to sense that I'm more used to handling data in Linux than Windows!If you made them on Linux, Notepad most likely won't do as it can't handle LF line endings (which is typical on Linux). Wordpad (or a third-party text editor) is required in this case. Last edited: 01/01/2015 at 23:22 by Muzer Log in to reply |
Info Dump 02/01/2015 at 02:38 #67292 | |
Danny252
![]() 1461 posts |
" said:" said:Ah, I thought I'd cut out all the initial blank files - apparently I was out by ten minutes.Yes, they're all just simple text files - Notepad or Wordpad will do. You might be able to sense that I'm more used to handling data in Linux than Windows!All good now. Turned out there were 2 blank files due to no trains in that time period, and they were the 2 I tried to open first. Quote: I've spent a bit of time in the last few days watching the CTA Train Tracker map so will match both my findings with your files, and see how they line up.I'd hope they come out the same, as both sets of data come from the CTA website! " said: If you made them on Linux, Notepad most likely won't do as it can't handle LF line endings (which is typical on Linux). Wordpad (or a third-party text editor) is required in this case.No Linux this time round, but the habit of naming any old data file ".dat" is a hard one to break. Log in to reply |
Info Dump 02/01/2015 at 03:17 #67293 | |
Muzer
![]() 718 posts |
They're lucky they even got ".dat" - I'm in the habit of not using any extension at all. Always a pain in the arse on Windows...
Log in to reply |
Info Dump 02/01/2015 at 13:08 #67300 | |
Danny252
![]() 1461 posts |
I tend to do that if it's just some notes or instructions that I'm scribbling down - but even that can be a bit painful for Windows!
Log in to reply |
Info Dump 11/01/2015 at 18:25 #67732 | |
guidomcc
![]() 246 posts |
I wrote a little program to take your data and produce little schedules for each run number. For example, train 519: Merchandise Mart 18:25:30 I've added what came out to the timetable spreadsheet. There are a couple of trains that got 'lost' during their run and I couldn't fit them in on the spreadsheet, noted here: #419, 0806 @ Merchandise Mart #414, 0848 @ Merchandise Mart #615, 1939 @ Roosevelt There are still some gaps. :huh: Last edited: 11/01/2015 at 21:16 by guidomcc Log in to reply |
Info Dump 11/03/2015 at 01:58 #69945 | |
puppypaws0123
![]() 1 posts |
" said:" said:Amazing work! :cheer:Are you making this into a SimSig WTT? Or is this data available for someone to use? :)That was my plan, but given that I'd seen two or three other people all say "I'll work on a timetable", I thought it better to contribute what I had to other's efforts! Log in to reply |