Network Rail timetable mileage data

[soil]

The Network Rail timetable contains distances between stations. Are these data made available in a machine-readable format anywhere?

 

[soil]

So I had a look at this.

Results attached in an Excel spreadsheet removed pending minor corrections.

There appear to be quite a few errors in the timetable.

Mileage for Manea is absent from the timetable.

Other stations missing include:
Shippea Hill
Lakenheath
Fishguard & Goodwick

There is a non-existent service listed - Stratford - Seven Sisters

There are some missing links:

Stratford - Tottenham Hale

Strawberry Hill - Teddington

Whitton - Hounslow

Chandlers Ford - Eastleigh

Cosham - Bedhampton

Slade Green - Crayford

Wishaw - Holytown - Motherwell

Hatton - Claverdon

The mileage for Claverdon also appears to be erroneous

Table 238 states that Dunbar is 10.25 miles from Drem. However interpolating from Table 26, the mileage should be 11.25 miles.

Haddenham & Thame Parkway reads 4 3/4, but should probably read 41 3/4

I have added in sensible values for the missing links/stations.

I have tried to exclude duplicates except where the fast and stopping services take a different route.

 

[FenMan]

This data is very useful and much thanks for making it available, however can I ask that this file (and the files posted by SickyNicky on another thread) is made available in a more portable format?

.xlsx is fine for those with recent versions of MS Office, but this file format doesn't always play nicely for other users e.g. SickyNicky's password-protected .xlsx file failed to open for me in Mac Office 2004, LibreOffice and iWork Numbers.

A .csv file would be accessible by pretty much all spreadsheet programs.

Many thanks!

 

[spasmj]

In case this is any use to you:

Station links file in format used by atoc (see http://www.atoc.org/clientfiles/File/RSPS5002 sp0037_12.pdf):

http://www.browngiant.com/RJRG0000.RGD

The data in the above file is derived from the pdf timetable so will contain errors.

 

[soil]

I wrote a quick web page that finds the shortest route by rail between two stations, plus other routes that are within three miles of this.

Maximum of six routes total:

http://www.bukitlawang.com/routes/routes.aspx

Still checking the data, there are probably a few minor errors.

You need to use correct station names - it won't attempt any kind of correction.

Here are some potential gotchas:

RICKMANSWORTH <LUL>
WHITECHAPEL <LUL>
WEST BROMPTON <LUL>
FARRINGDON <LUL>(LONDON)
CANADA WATER <LUL>
AMERSHAM <LUL>
CHORLEYWOOD <LUL>
HARROW-ON-THE-HILL <LUL>
CHALFONT & LATIMER <LUL>
GREENFORD <LUL>

Plus lots require a qualifier, such as:

CHARING (KENT)
HOPE (DERBYSHIRE)
MILLBROOK (BEDS)

etc.
--- old post above --- --- new post below ---
Ok so that seems to have crashed my entire web host server (Godaddy).

 

[John @ home]

Thanks a lot. It will be really helpful, particularly if we can iron out the glitches.

It worked well for the first two or three examples I set it, but hasn't responded since I asked it for the Worksop - Cleethorpes distance. This is difficult because the data appears not to have a distance for Retford - Gainsborough Central.

Do you want us to report problems like this in this thread?

 

[soil]

I am on the phone to Go Daddy about the server having crashed. They claim that it's a localised issue as apparently the hosting is clustered, so it should work from 'most of the world'. I am not sure if my page broke it, but I doubt it. They said it should be up 'real soon now'.

I checked Worksop to Cleethorpes on my local computer and it doesn't cause any error, suggesting that the shortest route is via Lincoln.

According to the National Rail map, the Retford - Gainsborough Central - Barnetby service runs only on Saturdays. Furthermore, this is not found in the eNRT timetable

While I am open to persuasion on the issue of 'regular train' or whatever the wording is, it does say that the mileage should be taken from the eNRT.

Therefore I would say that the shortest route from Retford to Barnetby is properly via Lincoln.

Although I say this, I did take the executive decision to add in some missing data, as detailed in my OP.

Also missing is New Beckenham - Beckenham Junction for the same reason (not in the timetable).

I am going to delete Stratford - Seven Sisters, though it is active currently.
--- old post above --- --- new post below ---
Edit: it seems to be up now.

 

[kieron]

I wrote a quick web page that finds the shortest route by rail between two stations, plus other routes that are within three miles of this.

It shows promise, but you seem to have omitted most of the links in central Liverpool.

The program omits the following routes shown on the NRT:
James Street -> Moorfields
Moorfields -> Liverpool Lime Street
Liverpool Lime Street -> Liverpool Central

It does include the following, which isn't shown on the NRT:
James Street -> Liverpool Central

You also seem to have included a link from Chesterfield to Belper. Belper is only listed on timetable 56, and this only includes stations to Matlock. The shortest route listed in the NRT goes via Derby.

While I appreciate that you can't simply use the distances in the NRT as they are (not least because they include links like Hartford <-> Warrington Bank Quay, which no trains follow), I would be interested to know how these particular differences arose.

Another thing I would find useful would be a list of the station names you use. It would probably be most useful to put a link to it on the web page.

 

[soil]

It shows promise, but you seem to have omitted most of the links in central Liverpool.

The program omits the following routes shown on the NRT:
James Street -> Moorfields
Moorfields -> Liverpool Lime Street
Liverpool Lime Street -> Liverpool Central

It does include the following, which isn't shown on the NRT:
James Street -> Liverpool Central

Thanks, I must confess to not being much of a Scouser, and so I didn't really think about the Liverpool loop.

I actually omitted most of the loop, but more generally I implemented all the links as undirectional (i.e. if you can travel from A to B you can also travel from B to A). The Liverpool loop looks to be the only place where this is not the case and trains all go the same direction..

I have fixed this now.

You also seem to have included a link from Chesterfield to Belper. Belper is only listed on timetable 56, and this only includes stations to Matlock. The shortest route listed in the NRT goes via Derby.

What stations did you use for this?

I tried it:

Shortest route from Belper to Chesterfield (32 miles)

Belper 0
Duffield 2.5
Derby 7.75
Chesterfield 32

Is this not correct?

Another thing I would find useful would be a list of the station names you use. It would probably be most useful to put a link to it on the web page.

I added an autocomplete now.
--- old post above --- --- new post below ---

In case this is any use to you:

Station links file in format used by atoc (see http://www.atoc.org/clientfiles/File/RSPS5002 sp0037_12.pdf):

http://www.browngiant.com/RJRG0000.RGD

The data in the above file is derived from the pdf timetable so will contain errors.

I had a look. For some reason Coulsdon South appears to have become Smithy Bridge (SMI).

I also note that the eNRT does not contain a mileage for Bristol Temple Meads - Taunton avoiding Weston-super-Mare.

For some reason the eNRT is around half-a-mile out on Rugeley - Lichfield. The distance for Stafford- Lichfield is 0.5 miles shorter (and correct), so I have added this in as an extra link.

Also Dundee - Broughty Ferry is given as both 59.25 to 63.25 and 83.75 to 87.25. The latter is correct, so I have corrected that.

I also added in Eccles - Manchester Oxford Road and Newton-le-Willows -Wigan North Western, as both are listed in the eNRT with mileage data, if infrequent.

I note that the mileages for Coseley are wrong.

Also Rugby - Coventry has two different values - I chose the lower (correct).

There is some confusion about Mitcham Junction - Hackbridge. This is down twice (tables 52 and 179) as 2 miles, once (table 182) as 0.75 miles. I have left it as 2 miles.

There are several errors in Table 43

For Scarborough - Hull Driffield is a palable error.

The positions of Filey and Hunmanby are different:

Scarborough - Hull:
Seamer - Filey 6.5 miles
Filey - Hunmanby 2.75 miles

Hull - Scarborough:
Hunmanby - Filey 3.75 miles
Filey - Seamer 5.5 miles

I have used the first table in Table 43, except I have taken the positions of Filey and Hunmanby from the second table.

 

[John @ home]

For some reason Coulsdon South appears to have become Smithy Bridge (SMI).

I think the code for Coulsdon South remains CDS.

• Stoats Nest for Coulsdon and Cane Hill (later Coulsdon North) was opened on 8 November 1899, by the LB&SCR on their Quarry Line which bypassed Redhill. It closed 3 October 1983.
• Smitham (later Coulsdon Town) was opened in 1904 by the SER on their Tattenham Corner Line and is named after another nearby settlement. This station was renamed from Smitham in 2011 as part of Southern Railway's new franchise agreement. This change was made as a result of a local consultation carried out by Croydon Council and it intended to better reflect the location of the station near Coulsdon town centre.
  
  http://en.wikipedia.org/wiki/Coulsdon_South_station

The code for Smitham was SMI until the 2011 re-naming as Coulsdon Town, which appears then to have taken the old Coulsdon North code of CDN, redundant since 1983. Code CDT (Caldicot) was unavailable. This change appears in the National Fares Manual but I am not surprised that it does not appear in other sources which are assumed to be permanent, such as mileage data.

The code for Smithy Bridge remains SMB.

 

[snail]

How is this any different to http://mileage.railmiles.org ?

 

[John @ home]

How is this any different to http://mileage.railmiles.org ?

railmiles uses Network Rail data about their lines, measured in miles and chains. It gives the shortest distances between measuring points, not just stations. Where the shortest distance includes reversal at a junction, it uses the distance to and from the junction only and does not mirror the typical passenger train's action of continuing to the next railway station.

The data supplied by soil is provided by the train companies' trade association ATOC and gives distances between railway stations, measured in units of quarter of a mile. This is the electronic form of the mileage data in the National Rail Timetable. The National Routeing Guide instructs us that "The shortest route is calculated by reference to the National Rail Timetable.".

 

[kieron]

Thanks, I must confess to not being much of a Scouser, and so I didn't really think about the Liverpool loop.

I actually omitted most of the loop, but more generally I implemented all the links as undirectional (i.e. if you can travel from A to B you can also travel from B to A). The Liverpool loop looks to be the only place where this is not the case and trains all go the same direction..

I have fixed this now.

Thank you.

I don't know if there's another table where the two halves don't mirror each other. What the tables do show is routes which are listed as being slightly longer in one direction. I don't know if that is ever significant, though.

There are other lines which could be used in either direction by scheduled passenger services, but aren't. One of the ones which is shown in the NRT is Wigan North Western => Newton-le-Willows, which gets 2tpd, southbound only.

This isn't something you can spot simply by parsing the NRT distance tables, though.

What stations did you use for this?

Sorry, I think I must have misread the results.

I added an autocomplete now.

I appreciate that, but that's only useful if you know how what characters each station starts with. There are a few stations on table 106 where the name includes a street reference. The only ones which also list the town are Liverpool Lime Street and Wallasey Grove Road. Other sources show different things.

How is this any different to http://mileage.railmiles.org ?

More simply, the shortest distance from Runcorn to Frodsham on railmiles is just over 4 miles. The shortest on bukitlawang is 34 miles, and you could probably walk there more quickly.

 

[soil]

Thank you.

I don't know if there's another table where the two halves don't mirror each other. What the tables do show is routes which are listed as being slightly longer in one direction. I don't know if that is ever significant, though.

I have been going through these and there are lots of errors in the timetable.

Here are some I found

Wigton, Aspatria - different on each timetable, I have chosen the ones that appear correct.

Glenrothes with Thornton - Kirkcaldy 7.25 vs 9 both incorrect should be 5. I have left it at 7.25

Whittlesea is one mile too close to Peterborough, but March is in the right place, so not a problem.

Sugar Loaf too close to Cynghordy in one direction

Nuneaton - Coleshill Parkway correct in table 49 but not 57. This is an absolute, not relative error (e.g., affecting the distance from Leicester - Birmingham New Street).

inverness - carrbridge correct in one direction only

Penkridge too close to Stafford in one timetable, too far in the other. Dudley Port ditto. I have left them both as they are internally consistent.

Larbert - Croy one mile different in each direction, corresponding with a one mile shift in bishopbriggs - Glasgow Queen Street

Worcester Shrub Hill - Droitwich Spa wrong in one direction.

Kirknewton, Currie Hill are in the wrong relative position on 3 out of 4 tables

The stations Wadhurst - Tunbridge Wells inclusive are in different positions in each direction.

Wilmcote - Hatton is correct in one direction only, on table 115.

Stansted Mountfitchet is in the wrong relative position in one direction only.

Dudley Port - Galton Bridge is shifted relatively in one direction only

Adlington - Leyland likewise.

Deptford - London Bridge is wrong in one direction by one mile.

I haven't considered any of the 'out by 1/4', which should just be rounding issues.

There are other lines which could be used in either direction by scheduled passenger services, but aren't. One of the ones which is shown in the NRT is Wigan North Western => Newton-le-Willows, which gets 2tpd, southbound only.

Yes I was aware of this.

I think there has been some discussion in the past on this issue, of whether the shortest route can include a train that runs only one way. I think the consensus was that it can. So perhaps I should make the Liverpool loop run in both directions! One or the other I guess.....

 

[bb21]

I think there has been some discussion in the past on this issue, of whether the shortest route can include a train that runs only one way. I think the consensus was that it can. So perhaps I should make the Liverpool loop run in both directions! One or the other I guess.....

If a certain member of ATOC staff's view is to be believed, the shortest distance can be different in different directions.

Not to say that it is a correct interpretation. I'll leave you to draw your own conclusions.

 

[soil]

If a certain member of ATOC staff's view is to be believed, the shortest distance can be different in different directions.

Not to say that it is a correct interpretation. I'll leave you to draw your own conclusions.

FWIW, I have now made the link in question southbound only.

 

[maniacmartin]

If a certain member of ATOC staff's view is to be believed, the shortest distance can be different in different directions.

Not to say that it is a correct interpretation. I'll leave you to draw your own conclusions.

I agree with this interpretation.

A friend of mine who used to work in IT at British Rail believes that the shortest route also varies depending on the day of the week, depending what services run that day. (A view that isn't commonly held I believe)

 

[kieron]

I have been going through these and there are lots of errors in the timetable.

I know. Some make no sense (such as Leamington Spa=>London Paddington being -3.25 miles via High Wycombe on table 115), some are inconsistent (such as Blackfriars=>Elephant & Castle being 1, 1.25 or -51.25 miles apart on table 52, depending on which column you check), and some are only an issue if you consider NRT distances to be subservient to physical reality (such as London Paddington=>Stratford-upon-Avon being 15.25 miles on table 115).

It is clear that ATOC don't use the distances for anything, and don't check the distances against anything they do use.

I think there has been some discussion in the past on this issue, of whether the shortest route can include a train that runs only one way.

If you allow that the shortest route can be a route you can't take using scheduled passenger services, then I'm fairly sure that there would be at least one railway ticket which isn't valid on any trains at all.

It's not Frodsham=>Runcorn, though, as they don't sell tickets for that route.

 

[soil]

It is clear that ATOC don't use the distances for anything, and don't check the distances against anything they do use.

True.

I have updated the route finder now to also give the RJIS shortest routes (using y/e 2011 data):

http://www.bukitlawang.com/routes/routes.aspx

I found the following links, which I have as one-way are implemented as two-way in RJIS:

LIVERPOOL CENTRAL LIVERPOOL LIME STREET
LIVERPOOL JAMES STREET LIVERPOOL CENTRAL
LIVERPOOL LIME STREET MOORFIELDS
LOCKERBIE CARSTAIRS
MOORFIELDS LIVERPOOL JAMES STREET
NEWTON-LE-WILLOWS WIGAN NORTH WESTERN
PATCHWAY BRISTOL PARKWAY
TOTTENHAM HALE CLAPTON
--- old post above --- --- new post below ---
Here's a list of the top paths in RJIS that are MUCH shorter than the timetable:

-160.45 LONDON KINGS CROSS LONDON MARYLEBONE 1.80 162.25
-158.41 LONDON CHARING CROSS LONDON MARYLEBONE 2.34 160.75
-151.48 LONDON MARYLEBONE LONDON VICTORIA 2.77 154.25
-147.26 LONDON MARYLEBONE LONDON PADDINGTON 0.74 148.00
-126.27 LONDON PADDINGTON SOUTH RUISLIP 10.23 136.50
-93.00 YEOVIL JUNCTION YEOVIL PEN MILL 2.00 95.00
-70.01 ACTON MAIN LINE KENSINGTON OLYMPIA 3.74 73.75
-63.39 CLITHEROE HELLIFIELD 13.36 76.75
-30.24 FRODSHAM RUNCORN 4.01 34.25
-26.60 ACTON BRIDGE (CHESHIRE) WARRINGTON BANK QUAY 9.65 36.25
-25.12 LONDON CANNON STREET LONDON FENCHURCH STREET 0.63 25.75
-24.94 DREM DUNBAR 11.31 36.25
-24.15 LONDON FENCHURCH STREET LONDON BRIDGE 0.85 25.00
-23.75 NEWPORT (S. WALES) ROGERSTONE 2.00 25.75
-22.60 STRATFORD (LONDON) STRATFORD INTERNATIONAL 0.40 23.00
-22.55 BERWICK-UPON-TWEED CHATHILL 20.95 43.50
-19.52 ELY KENNETT 14.73 34.25
-19.17 LONDON CHARING CROSS LONDON EUSTON 1.58 20.75
-18.45 WALSALL WOLVERHAMPTON 6.80 25.25
-17.92 BESCOT STADIUM WOLVERHAMPTON 6.08 24.00
-16.49 LONDON LIVERPOOL STREET LONDON BRIDGE 1.26 17.75
-16.03 LONDON EUSTON LONDON KINGS CROSS 0.47 16.50
-15.47 CARLUKE CARSTAIRS 8.28 23.75
-14.94 WATER ORTON WILNECOTE (STAFFS) 7.81 22.75
-14.55 LONDON FENCHURCH STREET LONDON LIVERPOOL STREET 0.70 15.25
-14.54 DALMENY LINLITHGOW 9.96 24.50
-12.45 LONDON WATERLOO LONDON WATERLOO EAST 0.05 12.50
-12.29 LONDON CANNON STREET MOORGATE 0.71 13.00
-11.84 LONDON EUSTON LONDON VICTORIA 2.41 14.25
-11.31 LONDON BRIDGE MOORGATE 0.94 12.25
-11.28 LONDON CANNON STREET OLD STREET 1.22 12.50
-11.25 BECKENHAM JUNCTION NEW BECKENHAM 0.50 11.75
-10.68 LONDON LIVERPOOL STREET MOORGATE 0.32 11.00
-10.32 LONDON KINGS CROSS LONDON LIVERPOOL STREET 2.18 12.50
-10.30 ST HELENS CENTRAL ST HELENS JUNCTION 1.95 12.25
-10.15 GUIDE BRIDGE STALYBRIDGE 2.10 12.25
-10.04 COATBRIDGE CENTRAL SPRINGBURN 7.71 17.75
-9.75 BURNLEY MANCHESTER ROAD ROSE GROVE 1.25 11.00
-9.60 LONDON EUSTON FINSBURY PARK 4.40 14.00
-9.41 LONDON BLACKFRIARS LONDON WATERLOO 1.34 10.75
-9.40 MARKINCH GLENROTHES WITH THORNTON 5.10 14.50
-9.20 DRONFIELD GRINDLEFORD 7.55 16.75
-8.72 MELTON MOWBRAY SILEBY 12.78 21.50
-8.67 LONDON CHARING CROSS LONDON VICTORIA 1.58 10.25
-8.66 MELTON MOWBRAY SYSTON 10.09 18.75
-8.60 LANGLEY MILL LOUGHBOROUGH 18.15 26.75
-8.55 DORE DRONFIELD 2.70 11.25
-8.52 ALFRETON LOUGHBOROUGH 24.48 33.00
-8.48 NEWARK CASTLE NEWARK NORTH GATE 1.52 10.00
-8.18 DALSTON KINGSLAND LONDON FIELDS 1.57 9.75
-7.85 LANGLEY MILL EAST MIDLANDS PARKWAY 13.15 21.00
-7.04 BIRMINGHAM NEW STREET SMALL HEATH 3.21 10.25
-7.02 CARDROSS HELENSBURGH UPPER 5.48 12.50
-6.89 PATRICROFT WIGAN NORTH WESTERN 16.11 23.00
-6.47 FOREST GATE WOODGRANGE PARK 0.28 6.75
-6.46 ANCASTER BOTTESFORD 12.54 19.00
-5.75 EAGLESCLIFFE YARM 2.50 8.25
-5.50 STRATFORD (LONDON) WEST HAM 1.00 6.50
-5.49 LONDON BLACKFRIARS LONDON VICTORIA 2.26 7.75
-4.97 LONDON VICTORIA LONDON WATERLOO 1.78 6.75
-4.26 RUGELEY TRENT VALLEY STONE 14.49 18.75
-4.14 BRIXTON DENMARK HILL 1.11 5.25
-4.14 LONDON KINGS CROSS MOORGATE 1.86 6.00
-3.86 CHISLEHURST ST MARY CRAY 2.39 6.25
-3.84 BRIXTON KENSINGTON OLYMPIA 5.41 9.25
-3.75 DEIGHTON BRIGHOUSE 3.25 7.00
-3.55 ASTON STECHFORD 2.95 6.50
-3.50 GOWERTON NEATH 11.50 15.00
3.45 OAKHAM CORBY 17.70 14.25
-3.41 BARNEHURST SLADE GREEN 1.59 5.00
-3.25 YATE CAM & DURSLEY 11.75 15.00
-2.94 NEWTON (STRATHCLYDE) UDDINGSTON 1.81 4.75
2.93 BIRMINGHAM NEW STREET KINGS NORTON 7.18 4.25
-2.85 SELBY YORK 18.40 21.25
-2.75 PEARTREE WILLINGTON 4.75 7.50
-2.75 BROMSGROVE ASHCHURCH FOR TEWKESBURY 24.00 26.75
-2.65 BROMSGROVE CHELTENHAM SPA 31.35 34.00
-2.55 DEAN GRATELEY 17.45 20.00
2.39 ACKLINGTON WIDDRINGTON 7.64 5.25
-2.35 BETHNAL GREEN STRATFORD (LONDON) 2.90 5.25
-2.31 ANCASTER GRANTHAM 9.19 11.50
-2.26 ACKLINGTON ALNMOUTH 3.99 6.25
-2.25 HEALD GREEN STYAL 1.50 3.75
-2.25 HEYSHAM PORT MORECAMBE 2.00 4.25
-2.21 EAGLESCLIFFE STOCKTON 3.04 5.25
-2.10 BANBURY KINGS SUTTON 1.40 3.50
2.00 LEICESTER SOUTH WIGSTON 3.75 1.75
-2.00 PORTLETHEN STONEHAVEN 6.25 8.25
-1.99 RUSKINGTON SPALDING 21.51 23.50
-1.95 ASHBURYS MANCHESTER VICTORIA 3.55 5.50
1.94 BICESTER NORTH KINGS SUTTON 12.44 10.50
-1.94 MANEA SHIPPEA HILL 13.81 15.75
-1.92 BURTON JOYCE LOWDHAM 2.33 4.25
1.89 ABERDEEN PORTLETHEN 9.89 8.00
-1.84 RUGBY WOLVERTON 30.16 32.00
-1.77 KILMARNOCK TROON 8.98 10.75
-1.75 STIRLING ALLOA 5.00 6.75
-1.67 ELSENHAM STANSTED AIRPORT 4.33 6.00
1.59 APPLEBY KIRKBY STEPHEN 12.34 10.75
-1.59 ATTENBOROUGH LOUGHBOROUGH 10.16 11.75

As you can see, they have implemented that the shortest path includes a walk. :roll: