478: quantum of sollazzo

#478: quantum of sollazzo – 26 July 2022

The data newsletter by @puntofisso.

Hello, regular readers and welcome new ones :) This is Quantum of Sollazzo, the newsletter about all things data. I am Giuseppe Sollazzo, or @puntofisso. I’ve been sending this newsletter since 2012 to be a summary of all the articles with or about data that captured my attention over the previous week. The newsletter is and will always (well, for as long as I can keep going!) be free, but you’re welcome to become a friend via the links below.

EU Twinnings, my good old hack from the EU Datathon 2020, became the subject of a data story on data.europa.eu. I’m happy about this. The idea behind EU Twinnings was this: can we create a measure of similarity between different areas? Eurostat data was quite handy, as it is relatively granular. There is quite a bit of missing data to complicate things, but the idea is a proof-of-concept that broadly works. In the past year or so my weekends have mostly been occupied by improving it thanks to a little commission that I gained from the EU Publications Office, but the concept is all there to be expanded upon, so any feedback is welcome. EU Twinnings is live here.

Every week I include a six-question interview with an inspiring data person. This week, I speak with Isaia Invernizzi of il Post. Isaia is a journalist who pioneered the use of data in local news, particularly getting his newspaper “L’Eco di Bergamo” to use data to cover the early, tragic phase of the COVID-19 pandemic in Lombardia. He’s now with the “Il Post” newspaper, which has a strong evidence-based ethos. Oh, and I shared a live discussion about data with Isaia on Italian national radio :)

The most clicked link last week, rather aptly given the heatwave, was this article from Bloomberg on how climate is pushing boundaries that are now beyond survivability for humans. Gloomy.

‘till next week,
Giuseppe @puntofisso

Six questions to...

Isaia Invernizzi

Isaia is a journalist at Il Post.

What is your daily data work like and what tools do you use?

I don't have a single workflow. I often start by searching for the latest data published by organizations or institutions to understand if there is a story behind it, or I ask myself a question and look for data that can give me an answer. I usually analyze them with Excel and visualize them with tools like Datawrapper, Fluorish or Tableau. I am increasingly using the data to find stories which I then follow with a reporting job, with field visits, interviewing people. It is truly inspiring and fascinating to use data in this way.

Tell me about a data project that you're proud of...

Surely my most important work was the analysis on over-mortality in the province of Bergamo during the early stages of the coronavirus epidemic. It was a very demanding job because at the time, at the beginning of March 2020, there was no reliable data to understand how many really died from Covid-19. We collected all the data from 243 municipalities to understand how many deaths there were compared to the average of the previous five years. This is a method that was then widely used in many other countries around the world, but which at the time had not yet been considered the only one to better understand the real dimensions of the tragedy.

...and a data project that someone else did and you're jealous of.

One of the works that I liked the most in recent years concerns the reconstruction of the collapse of the Morandi bridge in Genoa made by the New York Times in 2018. A team of journalists, engineers and technicians analyzed the videos of the collapse from which they extracted a lot of data to explain the causes of the accident. I would very much like to take a lot of time to work on such a large article in collaboration with technicians who can help journalists better tell the news.

If I say "dataset", you think of...

The first thing that comes to mind is that datasets often tell stories of people. In some ways, all analyses involve people or have an impact on them. For this reason, a lot of care is needed when analyzing datasets: it is often forgotten that what we think of as simple numbers actually have effects on people's lives.

Give someone new to data a tip or lesson you wish you'd learned earlier.

To ask ourselves many more questions: to find the data that are not there, that is missing, and also those that we do not realize we do not know. We must have an increasingly open and complete look at the news and data we need to work on a news or to tell a story. Often in the past I thought that having a dataset was everything and that it was enough to work on what we had in our hands, in fact I learned that analyzes cannot ignore the data that we lack.

Data is or data are...

Damn, you put me in a crisis. I believe that the wording "data is" is more correct, also because I think of data as a large world in constant motion.

Topical

London Is Set to Bake in 40-Degree Heat. Get Used to It

“Four charts show how more frequent heat waves in Britain are putting people at risk.“
Folks, we’ve got to do something about the climate and soon. Assuming it’s not too late already.

What’s the state of Southeast Asian sci-fi?

I found this article by Kontinentalist incredibly interesting because it starts from a question about something I now realise I had “felt” but never rationally thought about: sci-fi often presents urban futures that look Asian, but where there aren’t many Asian characters.

Covid-19 vaccines saved an estimated 20m lives during their first year

“Their impact in poor countries depends on how effectively governments prioritised recipients”. One of those great, clear charts and analysis from The Economist. (via Soph’s newsletter)

Who will be Britain’s next prime minister?

“Follow the contest between Liz Truss and Rishi Sunak with our interactive tracker”.

What’s Behind Biden’s Record-Low Approval Rating?

“After entering office, most recent presidents see a decline in their approval ratings, but since Biden’s numbers peaked at 55 percent in March 2021, they have moved in only one direction: down.“
FiveThirtyEight looks at the current president’s ratings.

‘Are the police capable of changing?’: Data on racial profiling in California shows the problem is only getting worse

The San Francisco Chronicle looks at stop-and-search data.

Tools & Tutorials

Project Naptha

An AI-driven Chrome extension to “highlight, copy, and translate text from any image”.

jupyter-scatter

Jupyter-scatter is an “interactive 2D scatter plot widget for Jupyter Lab and Notebook. Scales to millions of points!“

Clean up photos

“Remove unwanted objects from photos simply by dragging boxes.“
AI is useful for some things, then.

Social Bearing

This is a website that can extract “insights & analytics for tweets & timelines” for free. I’ve tried and it seems to work, but you’ll have to assess whether the information is useful.

Tidy Finance with R

“This website is the online version of Tidy Finance with R, a book currently under development and intended for eventual print release via Chapman & Hall/CRC. The book is the result of a joint effort of Christoph Scheuch, Stefan Voigt, and Patrick Weiss.“
Themes in the book are the classic parsing of financial data, asset pricing, modelling (including using machine learning), portfolio optimisation, etc.

Export comments

This tool will allow you to download all comments from a social media post to an Excel file.

We found the best JavaScript newsletter.
Bytes is probably the funniest web dev newsletter you’ll ever read (trust me). If you like our newsletter, I’ve got a feeling you’ll love Bytes too. There’s a reason 100k developers read it every week.

Data thinking

Coming up with research ideas

I’m cheating, as this article is not quite about data but about research in general. But the points it makes are easily adaptable to your next data project.

Measuring Gender Inclusion

“To reduce the barriers women-led businesses face in public procurement markets, we need to be able to identify women-led businesses and understand to what extent they are bidding on and winning government contracts. This includes an understanding of both the number and the value of the contracts, as well as their sector. We can take these measurements as a baseline and then implement measures with the goal of improving these metrics over time.“
My friends at SpendNetwork are doing some stellar work on highlighting data issues.

Dataviz, Data Analysis, & Interactive

The Rosetta Stone

Photogrammetry is great. Here’s a 3D model of the stone, originally created by the amazing Daniel Pett for the British Museum.

Ahmad Barclay on new census viz tool

Great Twitter thread by Six Questions graduate and ONS dataviz guru Ahmad Barclay, presenting his census population profile tool.

National Grid: Live Status

“The National Grid is Great Britain’s electricity transmission network, distributing the electrical power generated in England, Scotland, and Wales, and transferring energy between Great Britain and Ireland, France, the Netherlands, Belgium, and Norway. Most data comes from the Balancing Mechanism Reporting System and is updated at five minute intervals. Solar data comes from Sheffield Solar and is updated at half hour intervals.“
An energy dashboard.
(via Fabio Bertone)

How to do Twitter Network Analysis and Visualization in R

“*Using an example Twitter network of British politicians, learn how to do fascinating network analysis and visualization.

LUCAS Soil, the largest expandable soil dataset for Europe: a review

“The ‘Land Use/Cover Area frame statistical Survey Soil’ (LUCAS Soil) is an extensive and regular topsoil survey that is carried out across the European Union to derive policy-relevant statistics on the effect of land management on soil characteristics.“
Striking data and visualization.
(via Mark K Smitham)

Bird sounds

Visualized using AI. I’m not entirely sure why… but it comes with source code.

AI

The deep learning obesity crisis

AI models have become bigger and bigger.

What You Don’t Know You Don’t Know: Using Machine Learning to Unearth Audience Engagement Insights

“I’m excited to share a new development in how newsrooms can use machine learning to uncover and serve their communities’ hidden information needs.“

quantum of sollazzo is supported by ProofRed’s excellent proofreading. If you need high-quality copy editing or proofreading, head to http://proofred.co.uk. Oh, they also make really good explainer videos.

Sponsors* casperdcl and iterative.ai Jeff Wilson Fay Simcock Naomi Penfold

[*] this is for all $5+/months Github sponsors. If you are one of those and don’t appear here, please e-mail me