-
This has been way too long in the making and it's finally out:
"Spatiotemporal Model Cards", basically a cheat sheet for documenting #GeoAI models so that you'll thank your past self when you circle back to your docs in the future
π Markdown template ... https://github.com/anitagraser/model-card-template
π #OpenAccess paper with all the details ... https://dl.acm.org/doi/abs/10.1145/3809183Still planning to put the essential explanations into the template to make it easier to use without consulting the paper π©βπ»
-
Do I have any followers here who used scikit-mobility's human mobility metrics?
Since #skmob isn't maintained anymore, we're considering porting the essential functionality to @movingpandas and would love to have more feedback, especially if it's first-hand experience with the functionality:
-
@paleolimbot definitely ended up in the wrong dimension
-
Will we finally get nice html representations of @movingpandas Trajectories and TrajectoryCollections?
Inspired by #pydata #xarray π€©
WIP π©βπ» : https://github.com/movingpandas/movingpandas/issues/293
Any feedback / ideas welcome!
-
π°οΈ Geometry- and Physics-Aware Dataset Creation for Shadow Removal in High-Resolution Satellite Imagery https://meetingorganizer.copernicus.org/EGU26/EGU26-21279.html
π³ Establishing an Earth observation Super-resolution and Validation Framework for Improved Climate Hazard Assessment and Response in Forestry https://meetingorganizer.copernicus.org/EGU26/EGU26-21127.html
[2/2]
-
It's going to be a busy #EGU poster session for our team this year:
π² Explainable Machine Learning for diagnosing Data Quality Issues in Dendrometer-Based Tree Growth Time Series https://meetingorganizer.copernicus.org/EGU26/EGU26-21643.html
π Machine learning based dynamic numerical climate multi-model ensemble weighting for high impact weather affecting the energy infrastructure https://meetingorganizer.copernicus.org/EGU26/EGU26-18613.html
[1/2]
-
RE: https://mastodon.social/@xlth/116144192667591833
Not the ideal conditions for geospatial applications of VLMs π
#GIScience #VLM #spatiotemporal #MobilityDataScience #SpatialDataScience
-
Excellent video tutorial on creating animated traces in #QGIS over on #reddit:
https://www.reddit.com/r/QGIS/comments/1rde0fm/comment/o7c29l7/
This is just a sneak peak of the results:
-
Do I know anyone working with the #ECMWF? #FollowerPower #GISChat
-
I've added an interactive client script today. Makes it easier to play around and discover all the ways to break our little assistant:
-
QGIS to (Geo)Pandas β partΒ 3
The journey continues: QgsArrowIterator is now merged, making it possible to iterate overΒ QgsFeatures as Arrow batches.
This is where we are now ... http://anitagraser.com/2025/12/03/qgis-to-geopandas-part-3/
#QGIS #GeoPandas #GISChat #SpatialDataScience #Arrow #Python
-
The #AGIT2026 call for papers is out:
Short paper submission deadline: March 1, 2026
-
1st International Conference on Geospatial Artificial Intelligence (#GeoAI2026) will take place in Ghent, Belgium on 3-5 June 2026.
-
PhD Position: Spatio-temporal causal modeling of shared EV demand
100%, #ETH Zurich, fixed-term -
Getting a s***-ton of CSV files as input for spatial analysis can be a pain.
Yes, you can write a quick Python loop in the #QGIS console to load them but it's not a great workflow, imho.
The DuckDB read_csv function has been very convenient for this kind of ETL workflow:
-
Postdoc in the Spatial Data Science and GeoCommunication group at @univienna
https://jobs.univie.ac.at/job/University-Assistant-postdoctoral/1251977001/ -
What do you wish Data Scientists knew about spatial?
I'll be at #PyDataParis next week to spread the word
#GISChat #Python #DataScience #SpatialDataScience #PyData #Paris
-
The #MobilityDB team just announced v1.3.0-alpha featuring
New temporal types:
π tgeometry & tgeography that can represent the temporal evolution of any geometry type (polygon, multipoint, etc.)
π temporal circular buffer (tcbuffer)
π temporal pose (tpose) type, storing the evolution of a pose
(point position + orientation)and more ...
https://github.com/MobilityDB/MobilityDB/releases/tag/v1.3.0-alpha
-
Great discussions at #SDSL2025 about integrating #SpatialDataScience libraries in desktop and cloud environments, featuring @movingpandas , @qgis #Trajectools and the @carto Trajectory Analytics extension from the @emeraldseu project
-
π GeoAI Talks at UGent
The GeoAI Research Center (#UGent) warmly invites you to attend two consecutive #GeoAI Talks on Wednesday, 3 September 2025 (13:00β14:00).13:00 β Alex Levering (Vrije Universiteit Amsterdam)
Sensing landscape characteristics in a post-social media world13:30 β Ross Purves (University of Zurich)
Place, landscape and language
Join online via: https://lnkd.in/eqJ5JJ8t -
Interesting #MobilityDataAnalytics read by Molina-Toledo et al.: Negotiating space: The daily #mobility of women waste pickers in #Cuenca, #Ecuador
π https://www.sciencedirect.com/science/article/pii/S0966692325002698
Featuring #MobilityDB, #scikitmobility, @movingpandas, #trackintel and #dbt
-
#PostDoc position in #GIScience at @zgis.bsky.social
-
So happy to finally see Ana Basiri live, today at the #AGIT2025 Keynote
-
#DataSpaces panel at #AGIT2025, unfortunately ended up as a #manel
-
Thanks to all participants at our morning workshop on #PublicTransport and #AI!
Now everyone is gathering for the official #AGIT2025 opening session with many exciting new changes
-
If you're at #AGIT2025, I'd love to see you at our #MobilityDataScience workshop focused on #PublicTransport tomorrow morning π
09:00-10:15 Session 1A: WS: Mobility Data Science & KI β Potenziale fΓΌr den ΓΆffentlichen Verkehr
-
π€© #MobilityDataAnalytics & #GIScience all around:
Attending the @emeraldseu GA today. Presented progress on #Trajectools and our #explainableAI & #activeLearning for #MobilityDataScience. While simultaneously traveling to #AGIT2025 π
-
Our urban #MobilityAnalytics project EMERALDS hosts its next webinar tomorrow:
ποΈ 27 June | β° 14:30 CEST | πOnline
How The Hague is using data to manage urban risks.
Register π https://emeralds-horizon.eu/events/urban-dynamics-and-decision-support-risk-assessment-and-forecasting-public-events -
π New @movingpandas 0.22 and #Trajectools 2.6 just landed
Over 50% runtime reduction for many #MovementDataAnalytics tasks π
-
And here's @martinfleis et al.'s new street network generalization tool:
-
Another interesting new paper on street network analysis, HT @gboeing
π― it's surprisingly hard to count intersections
-
Direct links for convenience ...
π Paper: https://www.mdpi.com/2071-1050/17/8/3634
:github: Code: https://github.com/plus-mobilitylab/netascore/tree/v0.9.0
#Walkability #HumanMobility #MobilityDataScience #GIScience #GISchat #OSM
-
It's been a pleasure contributing to the "Spatial Data Science Languages" preprint spear-headed by @edzer together with @martinfleis @paleolimbot @robinlovelace @nowosad @josi @loreabad6
https://fosstodon.org/@martinfleis/114261790590024080 -
π₯³ New versions of @movingpandas and #Trajectools have just landed:
π’ππ²π¦ #MovingPandas 0.21.3 https://github.com/movingpandas/movingpandas/releases/tag/v0.21.3
:qgis: Trajectools 2.5 https://codeberg.org/movingpandas/trajectools/releases/tag/v2.5
π Please update your MovingPandas install to get all the improvements in #QGIS Trajectools https://plugins.qgis.org/plugins/processing_trajectory/#plugin-versions
#GISChat #MovementDataAnalytics #MobilityDataScience #Python
-
The quest for a fair TimeGPTΒ benchmark
At the end of yesterday's #TimeGPT for mobility post, we concluded that TimeGPT's trainingset probably included a copy of the popular BikeNYC timeseries dataset and that, therefore, we were not looking at a fair comparison ...
http://anitagraser.com/2025/03/29/the-quest-for-a-fair-timegpt-benchmark/
-
#TimeGPT for #mobility: Can foundation models outperform classic machine learning models for flowΒ predictions?
tldr; Maybe. Preliminary results certainly are impressive.
#MobilityDataScience #gischat #urbanmobility #geoai #foundationmodels
-
Only one more week to submit to ACM TSAS special issue on Urban Mobility https://dl.acm.org/pb-assets/static_journal_pages/tsas/pdf/TSAS_Special_Issue_Urban_Mobility-1737749682787.pdf
HT https://mastodon.social/@luis_de_sousa/113946946892555809
-
Today I'm at #edbticdt2025 to present MobiML, our latest contribution to facilitate #MachineLearning from movement data, building on @movingpandas, #pymeos, and others
More details:
https://github.com/movingpandas/mobiml -
The EMERALDS project @emeraldseu is hosting an MLOps webinar "EMERALDS Data Infrastructure and Development Frameworksβ on 21 February, at 11:30 CET, see https://emeralds-horizon.eu/events/emeralds-webinar-data-infrastructure-and-development-frameworks
The talks explore the design and implementation of a dedicated #MLOps platform built with specialised #mobility libraries and tools. This platform is tailored to support ML engineers in the development and deployment of machine learning models for real-world applications
-
-
#Moveapps is looking for a Technical Assistant (m/f/d) proficient in #RStats and/or #Python to improve their #MovementDataAnalytics plattform
-
I've had smoother releases π
But hey, #MovingPandas 0.21 is here https://github.com/movingpandas/movingpandas/releases
Make sure to update if you want to use the new #QGIS #Trajectools plugin version which will be released soon π
#MovementDataAnalytics #MobilityDataScience #Python #DataScience #GISChat
-
Updated my collection of open source projects in the #MobilityDataScience / #MovementDataAnalytics with new findings from #SDSL2024
https://github.com/anitagraser/movement-analysis-tools
Any more suggestions welcome
-
Leaving Prague with tons of new ideas and impressions.
Thanks @martinfleis and team for organizing and hosting this 2nd edition of SDSL. Looking forward to many more to come
#GISChat #SpatialDataScience #MovingPandas #MovementDataAnalytics #MobilityDataScience #GeoPandas #Opensource
-
The #QGIS Molusce plugin is really impressive: tons of downloads and lots of impact in the research community
Find out more about the latest developments in the latest #QGISOpenDay recording:
-
Solved the riddle of how to get #ChatGPT #DataAnalyst into a #Jupyter notebook π₯³
Made a #Geospatial Assistant that reads a zipped #Shapefile and performs #DataAnalysis
Supported outputs so far: text, image, and html
Blog post coming up soon π©βπ»
-
π Is anyone of you aware of a python solution for creating a bivariate map from a raster?