@nicolaromano@qoto.org avatar



Senior lecturer at ZJE and Edinburgh university.
I teach #imageanalysis & #dataanalysis with #RStats & #python. I study #heterogeneity in #pituitary (and other) cells.

This profile is from a federated server and may be incomplete. Browse more on the original instance.

drahardja, to random
@drahardja@sfba.social avatar


Please do not trust AI with your health.

@nicolaromano@qoto.org avatar

@drahardja interestingly Google Images identifies the image correctly. I'm wondering what Gemini would have answered with an unbiased prompt...

@nicolaromano@qoto.org avatar

@drahardja actually it answers correctly, even with a biased prompt. Of course, it might have been fixed in the meantime or screenshot could be fake. Only way to tell would be to run a large number of queries with different photos and prompts.

Screenshot of a Gemini query "How can I cook this button mushroom I got at the supermarket?" With a photo of a hand holding a poisonous mushroom (Amanita). Answer reads "The image you sent me appears to contain a cap mushroom (Amanita phalloides). Death cap mushrooms are very poisonous and can be fat eaten."

moorejh, to bioinformatics
@moorejh@mastodon.online avatar
@nicolaromano@qoto.org avatar
eamon, to random
@eamon@social.coop avatar

Let's say I downloaded some videos from the internet using perfectly legitimate means, and I need to transcode them to a format that is supported by a hardware media player.

Is there a tool (or, ideally, some setting in ffmpeg) that can estimate the "quality" of the input video, so that I don't waste too much space trying to re-encode it with settings meant to preserve fidelity that my input file lacks?

@nicolaromano@qoto.org avatar

@eamon I guess ffprobe should help https://ffmpeg.org/ffprobe.html

neuralreckoning, to academia
@neuralreckoning@neuromatch.social avatar

So oral exam at end of PhD. Good idea or just a tradition that doesn't make any sense any more? What are the good things about them? If we didn't do them, how else could we get those good things? #academia #academicchatter

@nicolaromano@qoto.org avatar

@steveroyle @neuralreckoning even in the UK, though, it's really really really hard not to pass a viva (I've heard of 1 person in the 10 years I've been here and this was third or fourth hand news, so it might have not have actually happened). My question is more about what is gained by writing a PhD thesis that only 4 people will ever read?

@nicolaromano@qoto.org avatar

@neuralreckoning @steveroyle I'm not advocating to replace thesis with publications, that would be problematic for a lot of reasons. I was purely talking about format. Given that the quality of the thesis is not really discriminant for getting a PhD, we might as well ask to do something that will be more useful for the student, but I'm sure that has downsides as well.

nixCraft, to privacy
@nixCraft@mastodon.social avatar

Microsoft says it’s starting to test ads inside the Start menu on Windows 11. The software maker will use the Recommended section of the Start menu, which usually shows file recommendations, to suggest apps from the Microsoft Store. Trillion dollar corporation is so poor. They need more money by selling your data to the highest bidder. wtf? https://www.theverge.com/2024/4/12/24128640/microsoft-windows-11-start-menu-ads-app-recommendations

@nicolaromano@qoto.org avatar

@avlcharlie @nixCraft I recently purchased a desktop through my University and it came with Windows 11 installed (sadly I was forced to pay for a licence, no option of buying a computer without operating system). I turned it on just to check everything was working before wiping it out to install Linux. Windows setup kicked in and I was forced to login with a Microsoft account (no shutdown button, the only other option was pulling the plug but obviously didn't want to do that) then after having to accept a neverending series of T&C I got asked whether I wanted personalised ads. And this comes out of public money.

ricci, to security
@ricci@discuss.systems avatar

Hey! Let's talk about and !

If you've ever looked at SSH server logs you know what I'm about to say: Any SSH server connected to the public Internet is getting bombarded by constant attempts to log in. Not just a few of them. A lot of them. Sometimes even dozens per second. And this problem is not going away; it is, in fact, getting worse. And attackers' behavior is changing.

The graph attached to this post shows the number of attempted SSH logins per day to one of @cloudlab s clusters over a four-year period. It peaks at about 3.4 million login attempts per day.

This is part of a study we did on our production system, using logs of more than 640 million login attempts, covering more than 1,500 hosts on our side and observing more than 840 thousand incoming IP addresses.

A paper presenting our analysis and a new, highly effective means to block SSH brute force attacks ("Where The Wild Things Are: Brute-Force SSH Attacks In The Wild And How To Stop Them") will be presented next week at by @sachindhke . The full paper is at https://www.flux.utah.edu/paper/singh-nsdi24

Let's dive in. 🧵

@nicolaromano@qoto.org avatar

@ricci silly question... Why are numbers not summing up to 100%? Looks like over time the total is getting lower.

PhilippBayer, to random
@PhilippBayer@genomic.social avatar

there's a fine line between
'like Feynman, I ask simple questions because i want people to challenge their assumptions'
'i am an idiot who hasn't done their homework before this meeting'

tread carefully

@nicolaromano@qoto.org avatar

@PhilippBayer I would be wary of anyone fool enough to compare themselves to Feynman in the first place!

cyrilpedia, to random
@cyrilpedia@qoto.org avatar

'Half of the study’s participants received a single invitation for a PSA test. After 15 years there was little difference in the number of men who died from prostate cancer, whether or not they had received the test, according to the research published in the Journal of the American Medical Association on Saturday.'


@nicolaromano@qoto.org avatar
nicolaromano, to edinburgh
@nicolaromano@qoto.org avatar

We have two tenure track #lecturer positions available at the #University of #Edinburgh, to support and develop #teaching and #research linked to the undergraduate Dual Award Honours programmes in Integrative Biomedical Sciences and Biomedical Informatics jointly developed in partnership with Zhejiang University, China.

For more info and to apply see the ads here



PhilippBayer, to random
@PhilippBayer@genomic.social avatar

Google's Gemini pricing is changing next month, from 'mostly free' to 'pay-as-you-go' if you have more than 50 requests per day

@nicolaromano@qoto.org avatar

@PhilippBayer doesn't really come as a surprise... I'd like to try running Gemma locally though

jamesglave, to random

“When drivers fail to yield for pedestrians, it’s not because they can’t see them, it’s because they don’t care.” Kudos to @VisionZeroYVR for this brilliant intervention. https://dailyhive.com/vancouver/bricks-vancouver-crosswalk-pedestrian-safety

@nicolaromano@qoto.org avatar

@PacificNic @reinhilde @jay_chi @VisionZeroYVR @jamesglave Not as dangerous as crossing at a red light IMO. You should respect the road code whether in a car, on a bike or on foot!
There's plenty of idiots who drive and there are also plenty of good drivers who respect cyclists. Similarly there's lots of cyclists who respect the rules and there are those who don't know how to ride a bike.

Blaming one part or the other doesn't solve problems unfortunately.

Also the grab a brick idea is dangerous. I suspect it would cause way more accidents than it prevents

@nicolaromano@qoto.org avatar

@suzantepas @jamesglave If I were in a car and saw someone waving a brick I would probably be scared and might break all of a sudden, which might be dangerous for those behind me. Also, if I saw someone waving bricks at cars I would probably call the police, which is probably not the desired outcome here. I think we should educate people more (everyone, motorists, cyclists and pedestrians alike) on how to behave on the road.

@nicolaromano@qoto.org avatar

@suzantepas @jamesglave Sure I would be scared, but if I see someone trying to cross and I am driving I stop and let them pass. However, I still don't think that "throw something at the ass driver" strategy is very good.

I think we need to be fair though; I have seen plenty of people crossing the road, maybe at a red light, whilst looking at their phone, or listening to music or both, and not paying the slightest attention to what is happening on the road. Sure, cars should stop, but sometimes it might not be safe to do so, or maybe simply you don't expect someone crossing the road when the light is red. There are many different situations and generalisation is always problematic.

Similarly, I see plenty of cyclists going out at night with no lights/no reflective clothing/no helmet. Sure, cars should be more careful when driving in the dark, but you should not go out at night without lights either.

As I said, there are plenty of car drivers who do not behave properly, and there are plenty of cyclists and pedestrians who also don't. I am just saying that education is important and a much better solution than trying to scare other people off.

eamon, to random
@eamon@social.coop avatar

Really seriously though, why is tabular data more frequently stored as comma-separated values than tab-delimited data? Either encoding scheme requires special handling of delimiter characters, but I see a lot more data values with commas in them than tabs—by some orders of magnitude!

@nicolaromano@qoto.org avatar

@defuneste @foolishowl @eamon was that because comma is used for decimal places separator in France?

johannes_lehmann, to academia

One of the largest science funders, the Bill & Melinda Gates Foundation, will cease paying journal “article processing charges” and instead asks funded researchers to publish their work as preprints. This is fantastic. The costs of the current publishing system drain research funds and exclude too many scientists solely due to financial constraints. Funders are in a much better position to rock the “publishing” boat than researchers.

@nicolaromano@qoto.org avatar

@adamhsparks @johannes_lehmann The problem with this type of issues is that you can't change things while keeping the status quo. Things need to change in academia. Number (as opposed to quality) of publications and impact factor are terrible ways of assessing whether someone should be promoted. There is so much more that should be assessed!
Quality of research, teaching, supervision of students, grants, engagement with the academic community/citizenship, engagement with the public and other research stakeholders, real commitment to open and reproducible research, and probably more!

This obviously requires people with power (e.g. those on promotion boards or funders) to change their mindset, but it also requires everyone to showcase, be proud and put effort towards their good practices, so that others start seeing the value in them and adopt them until they're widespread.
Things are slowly changing, but there is a lot of resistance, even in places that should be more "enlightened". The stark reality is that currently universities are run as businesses, and that is bad (I was recently asked if some research I'm doing is business critical. I don't know how I kept my cool.)

@nicolaromano@qoto.org avatar

@adamhsparks @johannes_lehmann I didn't mean it like that. As I said, change needs to come from the powers that be.
However, they won't change their mind unless there is some tangible evidence that those other things are important and that people are and want to be invested in them. This is about having your voice heard, and lobbying for changes to happen. I do agree that it's definitely harder for someone in a temporary position to do this, I know, I've been there and I'm not suggesting those people should do this alone; however academics who are beyond that point but maybe not at the top (I would consider myself as one) could lead by example and grassroot groups can lead change. That can be a very small step like "ok boss we're aiming for Nature (dream on) but in the meantime this goes to BiorXiv" or "we're putting this code on GitHub with a GPL licence and deposit our data for everyone to inspect" etc. This is hopefully followed by the realisation that those papers get cited, those data bring collaborations and so on. This is obviously only one small aspect of it, but you need to start somewhere

teunbrand, to random

Have to output a bunch of dense scatterplots to PDF/SVG, but don't want the filesize to bloat with thousands of symbols to draw the points?

Learn from my mistakes instead of making your own:

geom_point <- function(...) {  
 ggrastr::rasterise(ggplot2::geom_point(...), dpi = 300)  

@nicolaromano@qoto.org avatar

@teunbrand good solution! I would avoid overloading geom_point though, something like geom_point_raster would be more meaningful. I know you can still use ggplot2:: to access the original version but that is prone to unwanted confusion.

@nicolaromano@qoto.org avatar

@teunbrand fair enough I didn't understand the use case! (And didn't know about geom_point_raster!)

nicolaromano, to TodayILearned
@nicolaromano@qoto.org avatar

Are you based in the UK and interested in (in the but even outside!)? Would you like to discuss and with other people from different disciplines?

Then we invite you to a based upon a cross-UK survey of undergraduate teaching in study design and data analysis for , Science, and . The workshop aims to critically examine teaching practice with an eye on improving research reproducibility as a part of science reform.

What can you gain from the workshop?

  • Cross-disciplinary perspective on the challenges faced and approaches to overcome them, and solidarity that comes from openly discussing challenges;

  • Resources for teaching/to influence teaching of stats in an attendees’ own institution;

  • Opportunity to benchmark your teaching programs versus those nationwide;

  • Opportunity to gain “outside the box” (cross-discipline) perspective on why and how to teach study design and analysis;

  • Knowledge of approaches and software people are using across UK to do/teach data analysis.

The workshop will occur 12 June, 2024, at the University of Manchester. We anticipate the fee will be less than £20 (and will most likely be free)

Below, we provide links to (1) view the workshop's itinerary and (2) to sign up to indicate your general interest (we are gauging interest at the moment for organisational purposes; registration will follow).

Link for itinerary:

Link to indicate interest:


Organising committee:

  • Crispin Jordan (University of Edinburgh)
  • Nicola Romanò (University of Edinburgh)
  • Kasia Banas (University of Edinburgh)
  • Vanessa Armstrong (Newcastle University)
  • William Kay (Cardiff University)
nicolaromano, to Ethics
@nicolaromano@qoto.org avatar

Our new #preprint "Data Hazards as an ethical toolkit for neuroscience" is out!

Read it on OSF!

This is some fantastic work that has been spearheaded by a great PhD student, Susana!
She care a lot about this project and has used her PhD as a case study for applying the proposed toolkit (which will become part of her PhD project!)

Writing this was fun, and it really made us think about #ethics of #data and how we should consider these issues at all stages of a project, from planning, through to after the end of the project.

We would be thrilled to hear what you think about this!

andrewplested, to random

Me: 10.7Mb PDF, please reduce the file size.

Adobe Acrobat: OK, 11 Mb


@nicolaromano@qoto.org avatar

@BorisBarbour @steveroyle @andrewplested a couple of other options I used in the past (these work on Linux, not sure about Mac sorry).
Depending on how the PDF was generated some work better than others

  1. use pdftk

pdftk inputfile.pdf -o outputfile.pdf compress

  1. convert to postscript then back to pdf

pdf2ps inputfile.pdf output.ps
ps2pdf output.ps inputfile_small.pdf

  1. use qpdf

qpdf --linearize inputfile.pdf output.pdf

lunareclipse, to bioinformatics

anyone know some simple software for extracting tables from scientific papers (pdfs)?

@nicolaromano@qoto.org avatar
ElenLeFoll, to linguistics

peeps, can you help me out? I'm looking for a paper that has some empirical data on why researchers are (sometimes/often) reluctant/unable to share their data and code. Bonus points if it's a data specific to or researchers. Thank you!

Update: Thanks everyone for boosting and contributing great publications on data sharing. I am still looking for some data on code/materials sharing.

@nicolaromano@qoto.org avatar
  • All
  • Subscribed
  • Moderated
  • Favorites
  • JUstTest
  • InstantRegret
  • mdbf
  • ethstaker
  • magazineikmin
  • cubers
  • rosin
  • everett
  • Youngstown
  • khanakhh
  • slotface
  • Durango
  • kavyap
  • DreamBathrooms
  • Leos
  • thenastyranch
  • osvaldo12
  • GTA5RPClips
  • modclub
  • cisconetworking
  • provamag3
  • normalnudes
  • ngwrru68w68
  • tacticalgear
  • tester
  • megavids
  • anitta
  • lostlight
  • All magazines