got up at 5:30am to bash out a 40 mile ride to the Pennines (and stop for coffee half way) all before work. 10/10 idea

gutted to be leaving the young statisticians section of but it's really important that this job is passed around

it's been a blast - good luck friends

Relatively simple effort with frustratingly complex code to make it all work - how are the names of famous pairs trending over time within US election candidates?

Code: github.com/jcken95/tidytuesday

Happy Halloween! 🤡🎃🧟‍♂️

Gave my {creepr} package a hexsticker this year as a trick-or-treat!

A paper of mine was accepted last year and now the journal has moved publisher and hence style files. I do not want to have to fix this mess 😭

Big stonks for unless you're IBM! Had fun using corrr::networkwork_plot() for a really fast & effective way to identify correlations ... then spent ages colour coding the subtitle and adding the side plots

code: github.com/jcken95/tidytuesday

Minimalist for the Alone data (both graphically and programmatically - only libraries used were {tidyverse} and {showtext}!)

Code: github.com/jcken95/tidytuesday

A lot of data for this week's about bird sightings in 2020-2021 in the USA. Recycled some code from last week to produce a bar chart to show which habitats have the most birds

For bring your own data week I constructed a quick summary of my thesis - the 10 most commonly used words. "Approach" is the only non-technical word in this list, perhaps I should have looked for synonyms

Code: github.com/jcken95/tidytuesday

The code includes the script I used to generate the data, but I haven't included my thesis because I don't think my examiner will have it yet! I have however included a permuted version of my thesis, so you should be able to recreate the viz

did some actual statistics (!!) with this week. It doesn't seem to be very well known, but Pukelsheim's 3 sigma rule allows you to perform robust inferences with very few assumptions. Here it detects the impact of covid lockdowns and whatever else happened in 2021. The model is an AR(1) model which uses data up to time t to predict what happens at t+1. No predictions made for first 12 time points

Code: github.com/jcken95/tidytuesday

Paper on 3 sigma rule (paywall?): jstor.org/stable/2684253

so cold here right now that the _inside_ of the car is frozen

Wasn't really sure what to do with the elevator data for . I had recently come across {vapoRwave} for a e s t h e t i c themes and pallets, I thought I'd try plotting a retro video game style map

code: github.com/jcken95/tidytuesday

as it stands, my career is safe from AI and an AI can never specify your priors 🤷

Frustrating one for - struggled to get good spacing for the flags without mucking up the aspect ratio on the image of Bobby Moore.

Preserving aspect ratios for {ggimage} seems basically impossible. Trying to get good horizontal spacing warps the aspect ratio. Maybe with more time I could get it to work?

And although we beat (West) Germany in 1966, they've definitely punished us for it over the years ...

Code: github.com/jcken95/tidytuesday

the pencil museum isn't the only indoor attraction in the lake district! This week's looks at museums in the UK, decided I'd look close to home for this one. Had a bit of fun learning how to use {osmdata} drawing maps

code: github.com/jcken95/tidytuesday

Since this week's is about accessibility on the web, I decided to investigate options for making graphics more accessible

{ColourblindR} takes away the problem of choosing a CVD friendly palette by automatically generating a CVD friendly palette

Code: github.com/jcken95/tidytuesday

Blog post about {ColourblindR}: ubc-mds.github.io/ColourblindR

Show older
Qoto Mastodon

QOTO: Question Others to Teach Ourselves
An inclusive, Academic Freedom, instance
All cultures welcome.
Hate speech and harassment strictly forbidden.