Will Deakin / is on Nostr

I write code and visualize data
I do governance for British railway IT plus European regulation
I believe data makes rail passengers lives better
Ask me how!
#Physicist #OpenData #DataScience

Last Notes

2025-03-11 12:30:08

- reply

@nprofile…k2mh I've done it anyway. Can you read GeoPKG?

2025-03-11 12:03:10

- reply

@nprofile…k2mh the RDM is free for the data we're talking about but there are those would would happily charge you for data that is free elsewhere. I'll have a look now, although it might not be until this evening.
(The issue is that Cameron Bridge "only" opened in June 2024, so won't have necessarily made it to these report as they are 12-18 months after the fact).

2025-03-11 11:22:02

- reply

@nprofile…k2mh are you on the Rail Data Marketplace https://raildata.org.uk?
If not I can share the NaPTAN (National Passenger Transport Access Nodes) which has station locations and codes buried, or the ORR active station list?
If you're ok with a secondary source, have a look at point 4 and 6 here https://GitHub.com/anisotropi4/kingfisher?tab=readme-ov-file#data

2025-03-06 10:16:02

- reply

@nprofile…k2mh I have but it is feels a bit lame as it is about a building development blocking access to a pavement, and the council team getting pedestrian access fixed. Or reporting near misses by drivers whilst cycling.

2024-12-24 13:56:04

- reply

@nprofile…c5n8 ...or have multiple GB access points, for example a location South of the river adjacent to a Thameslink station might help...

2024-12-24 13:19:30

- reply

@nprofile…c5n8 the other issue is the whole security theatre at St Pancras and stations on mainland Europe. The current St Pancras layout is struggling with current passenger numbers at or near peak. Let alone when there is disruption. But you know this...

2024-12-24 13:13:02

- reply

@nprofile…c5n8 one of the economic challenges you way want to consider is around Eurotunnel track access charges. Which to my, albeit limited, knowledge is unique on the British Network for having a per occupied seat charge. Whereas all others are based on train service or, at least, on number of carriages.

2024-12-24 13:13:02

- reply

2024-11-29 21:17:44

For those interested in such things, following the publication of the Origin Destination Model by the Office of Rail and Road (ORR) last week, here is the six-year aggregated rail passenger flow for 2018-24 on an updated simplified centre-line track-model base model.
1/n
https://cdn.fosstodon.org/media_attachments/files/113/568/315/572/954/889/original/2be14173e1ba9cd9.png
https://cdn.fosstodon.org/media_attachments/files/113/568/238/335/406/546/original/3eedbf47cb8dd727.png
https://cdn.fosstodon.org/media_attachments/files/113/568/317/684/186/958/original/96bb791c9e020356.png
https://cdn.fosstodon.org/media_attachments/files/113/568/319/340/177/012/original/6646983e3ba7275a.png

2024-11-08 18:10:05

- reply

@npub194h…4ewx is this available on the internet?
#AskingForAFriend

2024-11-07 08:55:58

- reply

@npub1g0t…tvm4 I was talking to my son about this he quoted Gramsci "The old world is dying, the new slow to be born. It is in this stark contrast that monsters are born."
If correct, the market based neo-liberal economic model is dead. The question is, is the Labour government approach of competent management of the decline enough, or is a much radical say interventionist, planned, even Keynesian approach needed?
This is arguably where the US was in the '30s and the UK in '45.
1/n

2024-09-13 16:29:59

Is there an #emacs #Mastodon interface?
I ask as I think it is the one thing that would nail the vibe

2024-08-19 17:02:42

- reply

@npub1g0t…tvm4 please keep loving. We need more of this kind of thing.

2024-01-20 12:21:43

For those interested in such things I have had a go at writing a #python module "parenx" ("pare" + "nx") that simplifies linear networks, such as road and rail, using buffering and either image skeletonization or Voronoi polygons to identify a centre-line
As it is on #PyPi so it is available via #pip
It is beta-code and has a number of limitations but hopefully it might be of interest. Noting that two-dimension areal interpolation problem seems well understood
https://github.com/anisotropi4/parenx