Mistypes in the station names in the ORR data

[giacecco]

Unexpected, e.g. Handborough is in reality Hanborough. Heysham Harbour is actually called Heysham Port, etc. Should we change these manually if they aren't too many? Full list to follow...

[giacecco]

Lookup fails in the 5% of cases: 128 out of 2,526 stations listed in the ORR report.

.../orpi-corpus$ foreman run node main.js --out data/corpus-new.csv 
..............................................................................................................
Lat/lon resolution failed for Stewartby...............................................
Lat/lon resolution failed for Bedford St.Johns
Lat/lon resolution failed for Bedford Midland......................................................................................................................................................
Lat/lon resolution failed for Mossley Hill...........................
Lat/lon resolution failed for Bank Hall
Lat/lon resolution failed for Bootle Oriel Road....
Lat/lon resolution failed for Liverpool Lime Street.........
Lat/lon resolution failed for Old Roan................................
Lat/lon resolution failed for St.Helens Junction.........
Lat/lon resolution failed for Wigan North Western.............
Lat/lon resolution failed for Wigan Wallgate.....
Lat/lon resolution failed for Hope (Clwyd).........................................
Lat/lon resolution failed for Hall I' Th' Wood.....
Lat/lon resolution failed for Hag Fold.
Lat/lon resolution failed for Bolton
Lat/lon resolution failed for Brinnington..
Lat/lon resolution failed for Moses Gate..........
Lat/lon resolution failed for Littleborough................
Lat/lon resolution failed for Heysham Harbour.........................
Lat/lon resolution failed for Hazel Grove.
Lat/lon resolution failed for Davenport
Lat/lon resolution failed for Stockport.
Lat/lon resolution failed for Woodsmoor..
Lat/lon resolution failed for Bredbury.
Lat/lon resolution failed for Belle Vue
Lat/lon resolution failed for Flowery Field
Lat/lon resolution failed for Newton For Hyde
Lat/lon resolution failed for Ashton-Under-Lyne.....
Lat/lon resolution failed for Altrincham........
Lat/lon resolution failed for Hyde Central....
Lat/lon resolution failed for Hale....
Lat/lon resolution failed for Reddish South
Lat/lon resolution failed for Heald Green
Lat/lon resolution failed for Heaton Chapel
Lat/lon resolution failed for Levenshulme
Lat/lon resolution failed for Mauldeth Road.
Lat/lon resolution failed for Cheadle Hulme.
Lat/lon resolution failed for Rose Hill (Marple)...........
Lat/lon resolution failed for Rochdale.
Lat/lon resolution failed for Humphrey Park
Lat/lon resolution failed for Urmston.
Lat/lon resolution failed for Navigation Road
Lat/lon resolution failed for Ashburys
Lat/lon resolution failed for Bramhall
Lat/lon resolution failed for Burnage.
Lat/lon resolution failed for Chassen Road.
Lat/lon resolution failed for East Didsbury..
Lat/lon resolution failed for Gatley..
Lat/lon resolution failed for Ardwick
Lat/lon resolution failed for Manchester Airport
Lat/lon resolution failed for Gorton
Lat/lon resolution failed for Deansgate
Lat/lon resolution failed for Manchester Oxford Road
Lat/lon resolution failed for Manchester Piccadilly
Lat/lon resolution failed for Trafford Park
Lat/lon resolution failed for Manchester Victoria
Lat/lon resolution failed for Middlewood
Lat/lon resolution failed for Moston.
Lat/lon resolution failed for Ryder Brow
Lat/lon resolution failed for Reddish North.....
Lat/lon resolution failed for Handborough......................................................................
Lat/lon resolution failed for Clifton Down......
Lat/lon resolution failed for Lawrence Hill.
Lat/lon resolution failed for Bristol Temple Meads....
Lat/lon resolution failed for Parson Street.
Lat/lon resolution failed for Stapleton Road
Lat/lon resolution failed for St.Andrew's Road
Lat/lon resolution failed for Sea Mills................................
Lat/lon resolution failed for Bodmin Parkway
Lat/lon resolution failed for Bugle
Lat/lon resolution failed for Camborne
Lat/lon resolution failed for Falmouth Docks.
Lat/lon resolution failed for Lostwithiel
Lat/lon resolution failed for Luxulyan
Lat/lon resolution failed for Newquay
Lat/lon resolution failed for Par
Lat/lon resolution failed for Penryn.
Lat/lon resolution failed for Perranwell
Lat/lon resolution failed for Redruth
Lat/lon resolution failed for Quintrel Downs
Lat/lon resolution failed for Roche
Lat/lon resolution failed for St.Austell
Lat/lon resolution failed for St.Columb Road...
Lat/lon resolution failed for Truro.....
Lat/lon resolution failed for Calstock
Lat/lon resolution failed for Gunnislake
Lat/lon resolution failed for Causeland..
Lat/lon resolution failed for St.Keyne
Lat/lon resolution failed for Liskeard
Lat/lon resolution failed for Looe
Lat/lon resolution failed for Menheniot..
Lat/lon resolution failed for Sandplace
Lat/lon resolution failed for St.Germans
Lat/lon resolution failed for Saltash................................
Lat/lon resolution failed for Ynyswen.......
Lat/lon resolution failed for Pontyclun.
Lat/lon resolution failed for Llwynypia...
Lat/lon resolution failed for Mountain Ash.
Lat/lon resolution failed for Fernhill..
Lat/lon resolution failed for Ystrad Rhondda...
Lat/lon resolution failed for Taffs Well
Lat/lon resolution failed for Tonypandy
Lat/lon resolution failed for Trefforest Estate
Lat/lon resolution failed for Trefforest
Lat/lon resolution failed for Trehafod
Lat/lon resolution failed for Treherbert
Lat/lon resolution failed for Treorchy.
Lat/lon resolution failed for Whitchurch (South Glamorgan)
Lat/lon resolution failed for Ton Pentre..
Lat/lon resolution failed for Dinas (Mid Glamorgan)
Lat/lon resolution failed for Pontypridd.........
Lat/lon resolution failed for Cwmbach
Lat/lon resolution failed for Aberdare
Lat/lon resolution failed for Penrhiwceiber.....................
Lat/lon resolution failed for Johnston (Dyfed)........................................................................
Lat/lon resolution failed for Dovey Junction...................................................................................................................................................................................................................................
Lat/lon resolution failed for Shepherd's Well.......................................................................................................................................
Lat/lon resolution failed for Sutton (Surrey).........................................................................................................
Lat/lon resolution failed for Ryde Pier Head....................................................................................................................................................
Lat/lon resolution failed for Portsmouth Arms...........................................................................................................................................................................
Lat/lon resolution failed for Whitwell (Derbys)....................................................................................................................................................................................................................................................
Lat/lon resolution failed for Tees-Side Airport............................................................................................................
Lat/lon resolution failed for Bentley (S. Yorks).................................................................
Lat/lon resolution failed for Wavertree Technology Park....................................................
Lat/lon resolution failed for Falls Of Cruachan.....................
Lat/lon resolution failed for Rannoch For Kinloch Rannoch...............................................................................................
Lat/lon resolution failed for Prestwick Internat'nl Airport..................................................................................................................................................
Lat/lon resolution failed for Llanharan........................................
Completed. Success rate 95%.

[giacecco]

Added debug feature to script, e.g. you can do:

.../orpi-corpus$ foreman run node main.js --nominatim "Stewartby railway station"
err is null
latlon is {"lat":"52.0694881","lon":"-0.5204876"}

... and that shows that the stations failing actually resolve to good latlon! Hence this must be a sw bug or an issue in managing Nominatim's throttling. What is really odd is that it is always the same stations that fail.

G.

[giacecco]

Current output:

.../orpi-corpus$ foreman run node main.js --out data/corpus-new.csv 
...................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................
Lat/lon resolution failed for Heysham Harbour.......................................................................................................................
Lat/lon resolution failed for Handborough......................................................................................................................................
Lat/lon resolution failed for Quintrel Downs.........................................................................................
Lat/lon resolution failed for Whitchurch (South Glamorgan)......................................
Lat/lon resolution failed for Johnston (Dyfed)...............................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................
Lat/lon resolution failed for Whitwell (Derbys).................................................................................................................................................................................................................................................................................................................................................................
Lat/lon resolution failed for Bentley (S. Yorks)............................................................................................................................................
Lat/lon resolution failed for Rannoch For Kinloch Rannoch...............................................................................................
Lat/lon resolution failed for Prestwick Internat'nl Airport...........................................................................................................................................................................................
Completed. Success rate 100%.
Data written to data/corpus-new.csv.

[statzhero]

When / how do we input the missing stations manually?

[giacecco]

I had started a file with the fixes here https://github.com/theodi/orpi-corpus/blob/master/data/ORR-station-names-corrections-INCOMPLETE.csv . You may add to it when you find issues. I would add to this file only the minimum possible information with the fixes. I'll write the code to apply the fixes sometime soon.