Colin Gillespie's Avatar

Colin Gillespie

@csgillespie.bsky.social

Co-Founder of @jumpingrivers.com

128 Followers  |  200 Following  |  44 Posts  |  Joined: 17.11.2024  |  1.9246

Latest posts by csgillespie.bsky.social on Bluesky

I am pleased to announce that together with some friends, we are organising a workshop on Non-Reversible MCMC Sampling, taking place at Newcastle University from 8โ€“10 September 2025.

Details on the programme and registration can be found at the workshop website (sites.google.com/view/probai-...).

24.07.2025 11:28 โ€” ๐Ÿ‘ 30    ๐Ÿ” 12    ๐Ÿ’ฌ 2    ๐Ÿ“Œ 1

Ah but now you can use chatgpt to both summarise and generate them. Efficiency around...

23.07.2025 19:12 โ€” ๐Ÿ‘ 1    ๐Ÿ” 0    ๐Ÿ’ฌ 0    ๐Ÿ“Œ 0

I've had a few ai generated recruiter emails that watched and summarised my conference talks. It was pretty convincing, except I got a different one every day for three days

13.07.2025 09:11 โ€” ๐Ÿ‘ 1    ๐Ÿ” 0    ๐Ÿ’ฌ 0    ๐Ÿ“Œ 0

But the general "tech eco-system" needs to take some responsibility. Many other things require age verification, with the notable exception being "adult" websites. Even a little help to parents would be appreciated

11.07.2025 17:02 โ€” ๐Ÿ‘ 0    ๐Ÿ” 0    ๐Ÿ’ฌ 0    ๐Ÿ“Œ 0
Preview
An Introduction to Python Package Managers Python package managers are essential tools that help developers install, manage, and update external libraries or packages used in Python projects. These packages can contain reusable code, modules, and functions developed by other programmers, making it easier for developers to build applications without reinventing the wheel.

Python package managers are essential tools that help developers install, manage, and update external libraries or packages used in Python projects. These packages can contain reusable code, modules, and functions developed by other programmers.

#rstats #python #packagemanager

09.07.2025 10:44 โ€” ๐Ÿ‘ 4    ๐Ÿ” 2    ๐Ÿ’ฌ 0    ๐Ÿ“Œ 0
Blackadder election results
YouTube video by Nemanja Puriฤ‡ Blackadder election results

Accidentally of course youtu.be/h6mJw50OdZ4?...

07.07.2025 20:32 โ€” ๐Ÿ‘ 1    ๐Ÿ” 0    ๐Ÿ’ฌ 0    ๐Ÿ“Œ 0

Recursive - minus core/recommended

07.07.2025 09:11 โ€” ๐Ÿ‘ 1    ๐Ÿ” 0    ๐Ÿ’ฌ 1    ๐Ÿ“Œ 0

My gut feeling is there is survivorship bias. For older packages that aren't actively maintained, it's easier to stay on CRAN as you package needs updating less often - no dependencies being removed from CRAN or function changes.

07.07.2025 07:23 โ€” ๐Ÿ‘ 1    ๐Ÿ” 0    ๐Ÿ’ฌ 0    ๐Ÿ“Œ 0
Histogram showing the dependencies of packages on CRAN. The median is Q3.

Histogram showing the dependencies of packages on CRAN. The median is Q3.

As part of our #rstats validation, we scan the dependencies for all CRAN packages. Surprisingly, the median number is only 3 (the maximum is 49).

07.07.2025 07:23 โ€” ๐Ÿ‘ 7    ๐Ÿ” 2    ๐Ÿ’ฌ 3    ๐Ÿ“Œ 0

Based on audience feedback, they had a great time. Thankfully, the session wasn't recorded!

06.07.2025 17:25 โ€” ๐Ÿ‘ 3    ๐Ÿ” 0    ๐Ÿ’ฌ 0    ๐Ÿ“Œ 0
Preview
2025-07-03 PSI Can I trust this package? Colin Gillespie

If you are interested in the actual version, feel free to take a look docs.google.com/presentation...

06.07.2025 17:25 โ€” ๐Ÿ‘ 4    ๐Ÿ” 1    ๐Ÿ’ฌ 1    ๐Ÿ“Œ 0

Last week, I had lots of fun giving an #rstats talk at PSI where all images were removed - thanks to @mikeksmith.bsky.social's laptop. Since the final ten slides were screenshots, this was particularly interesting for the audience ๐Ÿคจ

06.07.2025 17:25 โ€” ๐Ÿ‘ 3    ๐Ÿ” 1    ๐Ÿ’ฌ 1    ๐Ÿ“Œ 0
Preview
Serving 200 million requests per day with a cgi-bin Jake Gold tests how well 90s-era CGI works today, using a Go + SQLIte CGI program running on a 16-thread AMD 3700X. Using CGI on modest hardware, itโ€™s possible to โ€ฆ

We ditched CGI in the late 1990s because of the overhead of starting, executing and stopping a process for every incoming request... turns out modern servers (plus languages like Go or Rust with a fast startup time) mean CGI isn't such a bad idea any more! simonwillison.net/2025/Jul/5/c...

05.07.2025 23:41 โ€” ๐Ÿ‘ 135    ๐Ÿ” 21    ๐Ÿ’ฌ 10    ๐Ÿ“Œ 3
Preview
R Package Quality: Documentation Not all R packages are clearly โ€œgoodโ€ or โ€œriskyโ€, most fall somewhere in between. This post introduces our scoring framework around package documentation. We investigate the different measures, then look at a few well known packages.

Not all R packages are clearly โ€œgoodโ€ or โ€œriskyโ€, most fall somewhere in between. This post introduces our scoring framework around package documentation. We investigate the different measures, then look at a few well known packages.


#rstats #packagevalidation #softwaredevelopment

03.07.2025 10:44 โ€” ๐Ÿ‘ 7    ๐Ÿ” 4    ๐Ÿ’ฌ 0    ๐Ÿ“Œ 0
Preview
Building Trust with Code: Validating Shiny Apps in Regulated Environments In regulated industries, Shiny apps must be reliable, reproducible, and well-documented. Validation builds trust, supports compliance, and prevents costly delays. Learn how to avoid common pitfalls and set up a lightweight, effective validation approach that scales with your needs. Validation isnโ€™t a blocker, itโ€™s a long-term advantage for you and your team.

In regulated industries, Shiny apps must be reliable, reproducible, and well-documented. Learn how to avoid common pitfalls and set up a lightweight, effective validation approach that scales with your needs.


#rstats #ShinyAppValidation #testing

01.07.2025 10:36 โ€” ๐Ÿ‘ 6    ๐Ÿ” 2    ๐Ÿ’ฌ 0    ๐Ÿ“Œ 1
Preview
Shiny in Production 2025: Lightning Talk Lineup We are pleased to announce the lightning talks for this year's Shiny in Production conference! In this blog post, we've pulled together all of the talk abstracts to give you a full view of what to expect!

We are happy to share that alongside nine full-length sessions, this yearโ€™s conference (8โ€“9 October) also features a lineup of fast-paced 5โ€‘minute lightning talks.


#ShinyInProduction #RStats #RShiny #ShinyApps #DataScience

30.06.2025 10:10 โ€” ๐Ÿ‘ 7    ๐Ÿ” 4    ๐Ÿ’ฌ 0    ๐Ÿ“Œ 0

So predictable that weโ€™ll see an explosion of digital products selling โ€œideas for million dollar businessesโ€ that you can โ€œjust vibe code quicklyโ€.

Basically: โ€œbuy my digital product for $500, spend $1,500 on Lovable / Claude Code and become a millionaire.โ€

Another hype train

30.06.2025 06:20 โ€” ๐Ÿ‘ 143    ๐Ÿ” 14    ๐Ÿ’ฌ 10    ๐Ÿ“Œ 3

I think the BBC is different. I get an actual bill from the BBC. So its more transparent and hence a target.

I've often wondered how popular vpns. Easy, but how many people use them.

An adblock is easy, but most people don't use it.

29.06.2025 14:24 โ€” ๐Ÿ‘ 1    ๐Ÿ” 0    ๐Ÿ’ฌ 1    ๐Ÿ“Œ 0

It politics. BBC is paid for by the UK license fee. Everyone with a tv pays ยฃ140. The right always wnet to reduce it, so there's pressure to find new income

29.06.2025 06:57 โ€” ๐Ÿ‘ 0    ๐Ÿ” 0    ๐Ÿ’ฌ 2    ๐Ÿ“Œ 0
Preview
R Package Quality: Package Popularity The popularity of an R package can be measured using package downloads and reverse dependencies. While these measures clearly have issues, they do provide a good indication of the package health. This post looks at these measures in more detail.

The popularity of an R package can be measured using package downloads and reverse dependencies. While these measures clearly have issues, they do provide a good indication of the package health. Our latest blog post looks at these measures in more detail.

#rstats #validation #litmus

26.06.2025 09:28 โ€” ๐Ÿ‘ 3    ๐Ÿ” 2    ๐Ÿ’ฌ 0    ๐Ÿ“Œ 0
Preview
Shiny in Production 2025: Workshops The details of all of our workshops for Shiny in Production are now live on the conference website. Read on for full details all in one place.

This yearโ€™s Shiny in Production Conference workshops cover the key aspects of building effective, production-ready Shiny apps: Testing for {shiny}, Asynchronous Shiny, Maps in Shiny, Figma & UI Design for Shiny.


#ShinyInProduction #RStats #RShiny #ShinyApps

24.06.2025 10:23 โ€” ๐Ÿ‘ 7    ๐Ÿ” 2    ๐Ÿ’ฌ 0    ๐Ÿ“Œ 0
Preview
R Package Quality: Validation and beyond! Not all R packages are clearly โ€œgoodโ€ or โ€œriskyโ€, most fall somewhere in between. This post introduces a scoring framework to help users assess package quality, based on documentation, code, maintenance, and popularity. We also share key principles to ensure the scores are useful, fair, and adaptable to different contexts.

At Jumping Rivers, we've developed a scoring framework to help users assess R package quality.

Our latest blog post, "R Package Quality: Validation and Beyond!", walks through this new framework and shares guiding principles that ensure the scores are fair, flexible, and context-aware.

#rstats #R

19.06.2025 13:15 โ€” ๐Ÿ‘ 8    ๐Ÿ” 4    ๐Ÿ’ฌ 0    ๐Ÿ“Œ 0
Preview
Shiny in Production 2025: Full Length Talks We are pleased to announce the full length talks for this year's Shiny in Production conference! In this blog post, we've pulled together all of the talk abstracts to give you a full view of what to expect!

We're happy to share the main talks for the Shiny in Production Conference 2025!

This year's lineup includes some great talks on using Shiny in real-world projects, from building apps to scaling them in production.

Looking forward to seeing everyone there!

#ShinyInProduction #RStats #DataScience

18.06.2025 11:32 โ€” ๐Ÿ‘ 11    ๐Ÿ” 4    ๐Ÿ’ฌ 0    ๐Ÿ“Œ 0
Preview
October 2025

Early bird registration for the Shiny in Production Conference closes at the end of the month, grab your ticket before then!

We'll be announcing the selected speakers in the coming days.


#ShinyInProduction #RStats #DataScience #Shiny

12.06.2025 10:53 โ€” ๐Ÿ‘ 5    ๐Ÿ” 2    ๐Ÿ’ฌ 0    ๐Ÿ“Œ 0
abstracts โ€“ R/Medicine 2025 Shiny is great, but what if you need more flexibility and performance? In this lightning talk, Iโ€™ll show how the reactR package helped me extend Shinyโ€™s capabilities to build ESQapp, a highly interactive and scalable application. By integrating React.js, I overcame Shinyโ€™s UI limitations, improved performance, and created custom components that wouldnโ€™t be possible with Shiny alone. If youโ€™re looking to level up your Shiny apps with modern web technologies, this talk is for you!

We're happy to share that our Data Scientist, Pedro Silva, will be speaking at R/Medicine this Thursday. Heโ€™ll be presenting on โ€œValidating Shiny Apps in Regulated Environmentsโ€, an important topic for anyone working with R in clinical or regulated settings.

#RMedicine2025 #RStats #ClinicalResearch

11.06.2025 12:35 โ€” ๐Ÿ‘ 2    ๐Ÿ” 2    ๐Ÿ’ฌ 0    ๐Ÿ“Œ 0

Sounds similar to @jumpingrivers.com except i was more of a coward and stayed part time in my other job. My wife is also a large part of the company as well!

10.06.2025 19:39 โ€” ๐Ÿ‘ 3    ๐Ÿ” 0    ๐Ÿ’ฌ 1    ๐Ÿ“Œ 0
Preview
Rethinking Image Formats Adding images to a web page used to be easy. Now with high-resolution screens and an array of modern image formats it can feel like much more work. This post explains why things have become more complex and suggestions solutions for making life a little easier and web pages more performant.

Our latest blog post is about dealing with images on web pages - a task that used to be straightforward but has been made more complex in recent years by high-density screens and new file-formats with varying levels of browser support.

05.06.2025 13:59 โ€” ๐Ÿ‘ 1    ๐Ÿ” 1    ๐Ÿ’ฌ 0    ๐Ÿ“Œ 0

Canโ€™t park there mate

02.06.2025 05:56 โ€” ๐Ÿ‘ 117    ๐Ÿ” 8    ๐Ÿ’ฌ 2    ๐Ÿ“Œ 1
Post image

UEFA CHAMPIONS LEAGUE - WE ARE BACK! ๐Ÿคฉ

25.05.2025 17:01 โ€” ๐Ÿ‘ 406    ๐Ÿ” 73    ๐Ÿ’ฌ 17    ๐Ÿ“Œ 31
Preview
Cheese rolling live: Woman, 20, says 'I don't like cheese' after winning her race Gloucester local Luke Briggs - dressed as Superman - earlier won one of the men's race, beating competitors in various outfits.

If the news cycle is wearing you down, take a look at the BBC reporting from the cheese rolling. "Woman, 20, says 'I don't like cheese'" www.bbc.co.uk/news/live/cd...

26.05.2025 12:59 โ€” ๐Ÿ‘ 25    ๐Ÿ” 1    ๐Ÿ’ฌ 3    ๐Ÿ“Œ 1

@csgillespie is following 20 prominent accounts