Opens profile photo
Follow
Robin Lovelace
@robinlovelace
Data scientist, transport modeller, geographer, programmer and researcher and into maps, bikes and energy.
Leedsrobinlovelace.netJoined October 2009

Robin Lovelace’s Tweets

Pinned Tweet
Really excited about this: for the next ~18 months I will be working with the Data Science team, , and others on the new 10DS Fellowship. Looking forward to working with Federico, Alex and everyone on data science in government 🚀
Quote Tweet
ADR UK & ESRC are pleased to announce the first No.10 Data Science (10DS) Fellows, Robin Lovelace & Federico Botta! They will collaborate with 10DS & ONS to design #research and produce #analysis to inform #policy. Read about the potential of this scheme: bit.ly/3HxJWzz
Image
33
25
224
Show this thread
#RStats/#Python crossover/collaboration in action. Both great languages that enable reproducible, transparent and accessible #DataScience with #opensource software. Opportunities for cross-pollination and inspiration in both directions 🚀
Quote Tweet
Hey #DataScience folks, enjoyed @hadleywickham & co's R 4 Data Science? Inspired by their excellent example, I'm putting together "Python 4 Data Science" here: aeturrell.github.io/python4DS/welc Definitely a work in progress but great to get any early user feedback! #rstats #python #pydata
7
15

Topics to follow

Sign up to get Tweets about the Topics you follow in your Home timeline.

Carousel

This is impressive and scary in equal measure!
Quote Tweet
I made an #EarthEngine app that uses #CellularAutomata and elevation data to simulate sea level changes. Try it out: aazuspan.users.earthengine.app/view/cellular- 🌊🏔️
Show this thread
Map of topography in the Pacific Northwest, showing water levels rapidly rising until only a few peaks are above water
GIF
1
3
Was super fun collaborating on this project! As Francisco says, (especially in the amazing new Visual Editor mode) + + = 🤩 4 collaborative, reproducible+ open manuscript writing. Be the change 🚀
Quote Tweet
Ahhh...and the great @robinlovelace introduced to a new (to me) reproducible writing set up which I am now keeping for the foreseeable future using @rstudio and latex. 💯 recommended. Link to the repo👉🏼 github.com/fcorowe/sim_chp
Show this thread
8
23
Great to see this out there #OpenAccess 🎉 Contains reproducible example using the new {simodels} #RStats package, test it out! If you're interested in modelling movement and spatial interaction models (SIMs), this could be the ideal place to start 🚀 #RStats #gischat
Image
Quote Tweet
Image
New preprint!!! This is a collaborative project w/ a dream team: @robinlovelace & @adam_dennett 🔥🔥🔥 It's a new chapter for a book led by another dream team @levijohnwolf, @ajheppenstall & @profrichharris 🔗doi.org/10.31219/osf.i 🧵
Show this thread
2
10
41
Show this thread
John Lindsay is a great speaker and outstanding developer. Met him at back in 2014 which inspired me to work on #opensource software in academia. This is highly recommended! #gischat #rspatial
Quote Tweet
Another excellent @MapScaping podcast with the legendary #WhiteboxTools creator Dr. John Lindsay! A #geospatial analysis platform with 500+ tools🚀📣 Twitter: @whiteboxgeo Website: whiteboxgeo.com GitHub: github.com/jblindsay/whit Podcast: podcasts.apple.com/us/podcast/whi
Show this thread
Image
2
3
13
Show this thread
This is magnificent - there's a lot for sf to catch up with now, to bring all this automatically to all users of sf.
Quote Tweet
Announcing 🥁🥁🥁🥁🥁 #rstats s2 🌏 version 1.1.0 on CRAN! Packed with new features, performance improvements, and completely rewritten internals to support implementations for other languages/runtimes. Blog post with all the goods here: dewey.dunnington.ca/post/2022/s2-v
Image
Image
Image
3
28
Want to use geographic data analysis to support + *influence* (more) sustainable planning policies nationwide 🌱? Could be ideal place for people with #RStats, #Python, #RSpatial, #geopy skills. Unique opportunity to use reproducible #DataScience 4 public benefit!
Quote Tweet
24 hours left to apply for two fantastic 'heads of data' jobs in the Spatial Data Unit @luhc on data capability/ infrastructure, and data strategy/ transformation. Details in the 🧵 below #DataJobs #DataScienceJobs #data #DataScience #LevellingUp
Show this thread
3
Same could be said of decarbonisation research in #academia: research fiddling at the margins for less than 1% gains is being overtaken by bold work by the likes of , and . And rightly so, respect+🙏4 tackling the #ClimateCrisis head-on 🌍🚀
Quote Tweet
With astonishing speed, the established environmental and conservation groups have been eclipsed by grassroots movements like @ExtinctionR, @Fridays4future, @GNDRising and @JustStop_Oil. Why? Because most failed to show the courage and leadership we need. theguardian.com/commentisfree/
Show this thread
1
2
7
This is a HUGE release, just read the blog post and there are loads of impressive changes here. Coolest one for me was reproducible example that highlights: most GIS software assumes that... THE WORLD IS FLAT (pretty sure this applies to etc ; ) so cool #gischat
Image
Image
Quote Tweet
Image
Image
Image
Announcing 🥁🥁🥁🥁🥁 #rstats s2 🌏 version 1.1.0 on CRAN! Packed with new features, performance improvements, and completely rewritten internals to support implementations for other languages/runtimes. Blog post with all the goods here: dewey.dunnington.ca/post/2022/s2-v
1
11
🚨🚨We have extended #Paper & session submission deadlines for #spatial #DataScience 2022 until 26th July🚨🚨 #SDSS22 will be online but we will have some local gatherings all around the world bc we all miss in-person #conferences (& drinks, of course!) bit.ly/3aKfS8r
Quote Tweet
We are pleased to announce the 3rd Spatial Data Science Symposium, September 22-23. The paper deadline is July 18 and session deadline is July 26. sdss2022.spatial-data-science.net #sdss2022 @STKO_UCSB @spatialUCSB @grantdmckenzie @AnahidBasiri @Joe_GISc Join us!
1
5
13
For anyone wanting to get stuck into exploratory data analysis and modelling, and learn to use visualisation as a powerful way to better understand phenomena revealed in datasets, using reproducible code, this is highly recommended!
Quote Tweet
Places are still available for both online and in-person study for our course with @rJBeecham Computational Methods for Social Data Science: Exploring Society Through Visualization and Modelling. Applications are open until Friday! essexsummerschool.com/summer-school-
Image
5
15
About time! Very glad that this dataset, which has huge potential to save lives, will finally be made open access 🎉 Many thanks , and to the companies who collect this *actual* (not legal) speed data, look forward to using it 4policy+research 🚀
Quote Tweet
Delighted to announce we've signed a contract with @OrdnanceSurvey to provide speed data for The Public Sector Geospatial Agreement (PSGA), allowing public-funded organisations to freely access this data from April 2023. bit.ly/3P8KwHq #geospatial #data #mapping
Image
2
6
39
Show this thread
Why Low Traffic Neighbourhoods are critical for reclaiming residential streets for *people*.
Quote Tweet
Research shows this is most likely not coincidence. In an attempt to shave seconds off its users' commutes, Google Maps is routing traffic through minor streets, once primarily used by local residents and delivery vehicles.
Show this thread
Image
1
8
35
A key point from the paper: modelling crashes on the road network makes sense because it allows aggregation + interpolation, + flagging dangerous route segments in need of intervention. For more on the paper itself, see this thread (and read it!):
Quote Tweet
Recent article by Andrea Gilardi, Jorge Mateu, Riccardo Borgoni, @robinlovelace adopting a hierarchical approach to model car crashes considering the street network layout 🔥 🔗Open access doi.org/10.1111/rssa.1 🧵
Show this thread
Image
1
7
Show this thread
The Welsh Government: currently planning to lower speed limits to 20mph (=32km/h) *in all built-up areas* across the country bbc.com/news/uk-wales- The German government: currently stopping 100+ cities that would want to do so from introducing a generalised 30km/h limit
4
14
56
Show this thread
In hindsight, may not have been the most constructive thing to retweet. We need more collaboration and understanding between different ecosystems and less polarisation. Next thing I put out on this will be about collaboration, it's possible + good:
Quote Tweet
Video is live here! youtube.com/watch?v=PbO5FV
Show this thread
3
Show this thread
Never thought I'd see mojitos as a criteria for deciding which language to use! Valid point in my experience tho: for data analysis, visualisation and modelling on regular (or #rspatial) datasets R just works, out-of-the-box ⚡️Fuel for #RStats/#Python/#JuliaLang debates 🔥
Quote Tweet
So sorry to all who I offended with a joke about R versus python. They are both fine, it’s just personal preference! Like, do you prefer to do a quick and correct data analysis and then sit by the beach sipping mojitos, or
Show this thread
2
2
7
Show this thread
This looks like a key study into the level of trust (and perceived safety?) that people walking and cycling have in autonomous vehicles vs conventional cars. Details below.. Interestingly: not much difference. Hypothesis: this is because ~10% of worst human drivers are so bad!
Image
Quote Tweet
We found in 3 common priority-based maneouvres (right turn into side road, overtaking a parked car, passing over a zebra) in simulation and the real-world that pedestrian, cyclist and driver trust in automated vehicles was similar. Other findings see: tandfonline.com/doi/epub/10.10
1
Key point: not all cycle (or indeed walking, wheeling or driving) infrastructure is alike 🛣️🚲 There are at least 6 different types of what many simply call 'cycleways'! Important to distinguish between them. Great illustration from just-published paper for future reference.
Quote Tweet
Image
Only 6% of on-road cycle lane length is physically segregated from motor vehicles (br image ask for area).We have also estimated, for the first time, compliance of on-road cycle lanes with UK national design guidance (LTN 1/20) in terms of cyclist separation from motor vehicles…
Show this thread
5
16
This is a big finding: while ~100% of roads are compliant with national guidance designed for motor vehicles, only ~20% of cycle lanes comply with national guidance *in London* (probably much lower outside London) according to analysis presented in this paper.
Quote Tweet
Just 20% of on-road cycle lanes were estimated as compliant. There is considerable variation between boroughs with Hammersmith and Fulham @LBHF having the most (65% of their cycle lanes compliant) and @harrow_council having the least (1%) @London_Cycling 4/5
Show this thread
Image
1
6
26
Show this thread
This is, to my knowledge, the most detailed #openaccess dataset on cycling infrastructure in the world 🌍 detailed description of it plus lots of beautiful visualisations made with reproducible #RStats code: just published 🎉
Quote Tweet
In it we demonstrate that there are spatial inequalities in cycling infrastructure between boroughs – infrastructure is generally located in Inner London – and these inequalities persist even we take population size and geographical area into account. 2/5
Show this thread
Image
12
32