GeoMariano (@Geo_mpm) header
GeoMariano (@Geo_mpm) profile picture

Geographer and Lead GeoData Scientist at @HomesEngland. Daddy of 2, Raspberry Pi Geek, and enthusiastic mountaineering. Also a Spaniard ( Valencia) in the UK.

England
Joined July 2012
Photo by GeoMariano (@Geo_mpm)Photo by GeoMariano (@Geo_mpm)Photo by GeoMariano (@Geo_mpm)Photo by GeoMariano (@Geo_mpm)Photo by GeoMariano (@Geo_mpm)Photo by GeoMariano (@Geo_mpm)Photo by GeoMariano (@Geo_mpm)Photo by GeoMariano (@Geo_mpm)Photo by GeoMariano (@Geo_mpm)Photo by GeoMariano (@Geo_mpm)
I wish more authors were like @tapoueh . He has updated his book #TheArtofPostgresql and for the people who bought previously the book we can get the new update for free from his website after a nice email. I have now some reading to do updating myself with #Postgresql18
25
A todos los candidatos a las elecciones. No puede ser que en 2026 para los que estamos en el extranjero sea tan difícil votar. De verdad no hay ningún medio tecnológico que evite que tenga que imprimir el voto y mandarlo en sobre pagando yo por el sello? Luego la gente no vota...
45
I remember a Geography professor when I was studying at Uni in Valencia saying that it doesn't exist the natural catastrophe, the reality is much simpler, humans have occupied those spaces, and if it wasn't for that, nature would continue using flood plains, deltas or marshland.
67
While transitioning to the world of Data Science from the Geospatial world, I find very hard to define hard boundaries where data science is not involved in any way, whether is stats, computer science, or just wrangling data. Anyone else with the same issues to try to no overlap?
60
DBSCAN (Density-Based Spatial Clustering of Applications with Noise) is an unsupervised learning algorithm that identifies clusters as regions of high data density rather than assuming a particular cluster shape. Given a radius ε and a minimum number of points MinPts, an observation x is considered a core point if its ε-neighborhood contains at least MinPts observations. Clusters are then formed by connecting density-reachable points, while isolated observations can be classified as noise. Unlike K-Means, DBSCAN does not require the number of clusters to be specified in advance and can discover clusters with highly non-convex geometries. Its behavior is governed by the local density structure of the data, making it closely related to ideas from nonparametric statistics and density estimation. DBSCAN has applications in statistical machine learning including spatial data analysis, anomaly detection, image segmentation, customer behavior analysis and geospatial pattern recognition. From a statistical perspective, its notion of local density provides a simple way to distinguish structured populations from outliers. It is particularly useful when clusters have irregular shapes or when observations that do not belong to any meaningful population should be explicitly identified as noise. More recent density-based methods, such as HDBSCAN, extend these ideas to datasets with varying density levels.
26
1
185
7,468
Devastating content but top tier cartography by José Luengo-Cabrera’s highlighting the power of mapping. 𝐌𝐨𝐫𝐞 𝐟𝐫𝐨𝐦 𝐭𝐡𝐞 𝐚𝐮𝐭𝐡𝐨𝐫: linkedin.com/posts/jos%C3%A9… 𝐌𝐨𝐫𝐞 𝐠𝐞𝐨𝐬𝐩𝐚𝐭𝐢𝐚𝐥: thenewscienceofmaps.com/?utm…
26
195
13
715
66,713
My first impressions about #disctodigital from @XBOX So far I've been able to claim most of my Xbox games such as Assetto Corsa or NFS Heat. However none of the Forza version at the moment, nor the crew motorfest. Also none of the Xbox360 or original Xbox were compatible
97
GeoMariano retweeted
`now()` is not now, it's more of a was. `now()` is actually the start of the current transaction. The wall clock continues to move. Every later `now()`, `CURRENT_TIMESTAMP`, `CURRENT_TIME`, and `CURRENT_DATE` in a transaction still returns same time. Postgres calls this a transaction-frozen timestamp. A 10,000-row `INSERT` with `created_at DEFAULT now()` gets one timestamp, not 10,000 slightly different ones. Three clocks • `now()` is a Postgres tradition (among other databases) and is not a SQL-standard. The SQL standard is `CURRENT_TIMESTAMP`. (Additionally, Postgres has the `transaction_timestamp()`method is is the same as `now()`, but with a name that admits what it is.) • `statement_timestamp()` moves once per statement, which is useful inside a long transaction when you want per-command times without going all the way to the wall. • `clock_timestamp()` is the only `timestamptz` that changes during a single statement
6
11
136
8,007
⭐️⭐️ ¡¡𝗦𝗢𝗠𝗢𝗦 𝗕𝗜𝗖𝗔𝗠𝗣𝗘𝗢𝗡𝗘𝗦 𝗗𝗘𝗟 𝗠𝗨𝗡𝗗𝗢!! #VamosEspaña | #CopaMundialFIFA
6,009
89,934
5,583
266,124
11,915,087
⭐⭐ Proud again of the boys and their way of understanding football. I would have never imagined that I would live to celebrate 2 World Cup titles.16 years ago saw something amazing,today another extraordinary match and congratulations to the players of Spain footbal team #vamos
1
150
GeoMariano retweeted
Google set out to tackle the hedgerow problem. The hedgerows, stone walls and little copses doing much of England's ecological work are too small for any national forest map to see, so on paper they barely exist. Google decided to fix this. National forest inventories only count woodland above a certain size, so the thin ribbons of trees along field edges, the shelterbelts, the lone stands in a pasture all fall through the cracks, even though together they store carbon, shelter wildlife and stitch fragmented habitats back together. Google has now mapped every one of these features across the whole of England and turned them into data you can actually work with. Start with the scale of it. England covers more than 130,000 square kilometres, and the system draws the fine woody features across all of it. The surprising part is how little hand-labelled data that took: roughly 247 square kilometres of carefully annotated samples, a rounding error against the national area. The model could stretch that far because it sits on a foundation trained on more than 300 million satellite images from around the world, so it arrived already knowing what trees, edges and fields tend to look like from above, and needed only a little local teaching to specialise. To separate features that overlap and sit at different heights, it reads two layers at once: submetre optical imagery for the fine shape, and one-metre LiDAR, which measures height from the air, to tell a low hedge from the tree standing behind it. Then it sorts each feature by what it does in the landscape, using a simple measure of shape. Round, chunky patches above thirty metres across count as woodland. Long, thin, winding shapes get flagged as linear features like hedgerows. Shape becomes a stand-in for ecological function, worked out automatically across thousands of map tiles. The whole national sweep ran in parallel on Google Earth Engine, chewing through thousands of satellite cells at once rather than grinding through the country one region at a time. The new release, Vectorized Farmscapes 2020, turns that into clean geometric shapes, each hedgerow or copse its own object with a length, an area and a position. You can measure it, add it up, and track whether it grew or vanished between surveys. link to full article: research.google/blog/from-pi…
2
14
1
104
9,938
Transform document images into structured data with LlamaParse (automated validation) 📊 Converting document images such as receipts to structured spreadsheet data requires tedious typing and careful validation. LlamaParse automates document data extraction by combining OCR parsing with schema validation, eliminating manual typing and human error. Here is an example pipeline for extracting receipt data: • Parse receipt images to markdown using LlamaParse OCR engine • Define receipt structure with Pydantic models (company, date, items, totals) • Extract structured data automatically with OpenAI integration • Validate types and enforce business rules (positive prices, valid dates) • Export to pandas DataFrames or spreadsheets for analysis 🚀 Full article: bit.ly/46ODSl3 ☕️ Run this code: bit.ly/46SQrdF #DataScience #Python #MachineLearning #AI
3
15
733
GeoMariano retweeted
I’ve been deep in R for GIS mapping for years, but today marks a thrilling pivot: my very first Python tutorial is live! We’ll extract Google’s latest satellite embeddings, run k-means analysis, and map our results Everyone, watch the tutorial 👉youtu.be/WjoB7mou2n8
2
10
95
4,091
GeoMariano retweeted
Remember Global Buildings Atlas with 2.75 BILLION footprints for 97% of world's buildings? You can view buildings for any city using this interactive map. Link below 👇
2
22
2
161
11,305
THREAD: 1/4 If you have a retro console that has broken I would like to honestly recommend @topvint . My OG Xbox stopped working and didn't even switch. I contacted his business and quickly put my mind at ease thinking it would be a capacitor but would need to check.
2
1
7
765
3/4 With sometimes a lot of cowboys out there charging ridiculous prices just to open an Xbox, finding that #Topvint actually charges a normal price and actually cares about the repair was a blessing. Also super thorough with the repair and diagnostic
1
90
4/4 So if you are in need for a #RetroConsole repair or for some modding too, that is the place to go at @topvint . Very friendly but also professional, and for people like me that treasure my old consoles, trust is very important.Hope they get the credit they deserve #recommend
2
91
Anybody else prefers the streamlit folium library to the native @streamlit pydeck? I've found more customisation and easy to render maps with folium. But I would like to hear other people's thoughts. Labels and widget controllers are a win in Folium, but 3D render nice in pydeck
96
So awesome to have DeepMind 🧠 helping create datasets for #EarthEngine! Check out the Satellite Embedding dataset created with the AlphaEarth Foundations model 🛰️🌍👇 medium.com/google-earth/ai-p…
Our new AI model AlphaEarth Foundations is mapping the planet in astonishing detail. 🌏🔍 Scientists will now be able to track the impact of deforestation, monitoring crop health, and more – significantly faster, thanks to our new datasets. 🧵
7
43
2,449