Oscar Kjell

1/5 Did you ever wish to make a visualization of language data? 🤔

We made an R-based tutorial for social scientists on how to turn language data into visual insights ✅

Preprint: doi.org/10.31234/osf...

the R-packages text 1.6 and topics .6 are now on CRAN

topics includes visualisation improvements

text is now easier to install, it checks for dependencies and instructs how to install them

hope to get more user feedback to improve installation further

www.r-text.org
#rtext #rtopics #LLM #AI

hi again, please note that text 1.6 and topics 0.6 is now on CRAN.... and this version of text does not require the user to download JAVA. The installation is now made easier by explicitly checking and providing instructions regarding system dependencies. (again, thanks for the feedback)

HiTOP Teaching Hub Information about HiTOP self-report measures, including HiTOP-PRO, the harmful substance use module, and B-HiTOP

We’ve just launched the HiTOP Teaching Hub—a space to collect and share materials for teaching dimensional models.

If you have anything you're willing to share, please contact @kathleenwreardon.bsky.social or @miriforbes.bsky.social

Here's what we have so far: www.hitop-system.org/hitop-teachi...

The LEADING reporting guideline is now published in Comprehensive Psychiatry 🥳

See www.sciencedirect.com/science/arti...

Or read the previous summary 👇

Erroneous version of the chart with a significant share of cancer deaths being from lymphatic and blood cancers, across all ages.

Fixed version of the chart with a much smaller share of cancer deaths being from lymphatic and blood cancers, especially smaller during childhood.

1/ Correction! 🚨 I found and fixed an important coding error in a chart I made.

Specifically, the peach category 'lymphatic and blood cancers' in this chart wrongly included both lymphoid and blood cancers (C81-C96) plus some of its subcategories, meaning some were double-counted.

Before & after:

1. which text version are you using? have you installed the latest version from CRAN 1.5?
2. when you try this, can you please first always remove the embeddings that are automatically saved in your working directory (called something like textPredict_694....RDS).

i'm not able to reproduce this (some colleagues have tried too). if you run it in a standard R script, do u get some more info? (every time you try, pls use different texts input so u make a new embeddings every time). im also happy to schedule an online meeting next week to try n sort this out :)

I’m thrilled about this special issue — please submit your abstract!

Today’s reading exercise.

if that does not work, can you please try running this:

res <- textAssess(
model_info = "valence_facebook_mxbai23_eijsbroek2024",
texts = "test one")

in the console it says you are loading some embedings called textPredict_..... These are created automatically when running textPredict() with a new text. Now these embeddings are named differently; so try setting dim_name = FALSE within textPredict call.

Thanks for your patience. Do u already have a word embeddings for this text in your working directory?
Can u pls 1) try with a different text, or 2) delete any word embeddings created from textPredict() in ur working directory.
(i should add improve the warning!).

yes, thanks for letting me know. i've been thinking about this too. there are a few functions that i would prefer not having in both packaged to simplify maintenance. but, i will look into a solution so that the text-package installation does not require rJava. ill keep you posted.

have you tried to first downloading java from www.java.com/en/download/.

hmm what is the error? Have you tried installing rJava first?
here you can see more details about installing the text package (r-text.org/articles/hug...)
im happy to get feedback if it does not work.

Version 1.5 of the #r-text package is now available from #CRAN.

This version improves the application of pre-trained language assessments from the #LBAM-library (r-text.org/articles/LBA...).

#mlsky #PsychSciSky #Statistics #PsychSciSky #StatsSky #NLP

Differentiating balance and harmony through natural language analysis: a cross-national exploration of two understudied wellbeing-related concepts Although the concepts of balance and harmony are increasingly appreciated as important in research on wellbeing, their precise meaning is often vague or unclear. This paper aims to elucidate these ...

1/8
Balance ⚖️ vs. Harmony ☯️: Two emerging well-being constructs often sloppily used interchangeably.
But they’re distinct.

Balance = broad life domains ⚖️
Harmony = relationships🧘

We asked 15,275 people across 154 nations to show it.

www.tandfonline.com/doi/full/10....

Thanks for tagging me 🙌 looks really interesting? Would you like to present this to our research lab in an online meeting?

At least one Doctoral student in Psychology The Department of Psychology is now looking for a doctoral student in any field of study in the subject and based on their own proposed project plan. Regarding the doctoral programme The doctoral prog

our department at Lund University has a position for a PhD student.

feel free to reach out if you are interested in applying.

lu.varbi.com/en/what:job/...

tidymodels Internship for 2025 - Tidyverse The tidymodels team is sponsoring a summer internship in 2025.

The tidymodels group has an #rstats summer internship position this year.

The focus is on feature selection tools. All of the information is in the blog post below.

We look forward to seeing your applications!

www.tidyverse.org/blog/2025/01...

Thanks for the info! I will check it out!

I really wish I discovered the {S7} and {cli} package before -- it would have really streamlined so much package development

100% recommend these tools for any package developers.

#rstats

How come? Any advice on where to start learning more? 😁

A NASA photo of NGC 2264- a cosmic tree that has been named the Christmas Tree Cluster which lives in our galaxy about 2,500 light-years from Earth. It contains baby stars that are between 1 and 5 million years old 🥺

when the star on the tree is an actual star lol

A NASA photo of the Christmas Tree Cluster 🎄

Christmas card of a small tree, wrapped gifts and a cat.

We received this beautiful handcrafted Christmas card from a student from Germany who visited us earlier this year... she did an amazing job capturing our cat! 🎄⭐

Merry Christmas, everyone!

Version 1.3 of the #r-text package is now available from #CRAN.

This new version makes it easier to apply pre-trained language assessments from the #LBAM-library (r-text.org/articles/LBA...).

#mlsky #PsychSciSky #Statistics #PsychSciSky #StatsSky #NLP

Create searchable Bluesky bookmarks with R How to save and search all the Bluesky posts you’ve liked with the atrrr R package and use NotebookLM to ask natural language questions about those likes.

Better than bookmarks? Use R to save all your Bluesky likes in a searchable table!
Bonus if you use generative AI: Upload them to a free service like Google NotebookLM and ask natural language questions to find that post you remember.

www.infoworld.com/article/3623...

#Rstats #R #GenAI #Bluesky

A hex logo with 'lintr' written in light blue stitch. The angled white band behind this lettering appears to be "wiped on" by a lint roller at the right end of the hex. The rest of the hex is a Nordic knit sweater pattern in various shades of blue.

#rstats {lintr} finally got a hex logo! Thanks to @gregswinehart.bsky.social! And of course @hadley.nz for the initiative.

(PDF) Tell Us What You Really Think: A Think Aloud Protocol Analysis of the Verbal Cognitive Reflect... PDF | The standard interpretation of cognitive reflection tests assumes that correct responses are reflective and lured responses are unreflective.... | Find, read and cite all the research you need o...

👋 I’m Nick.

I study how we can make better decisions.

Some examples:
- Bounded Reflectivism & Epistemic Identity: researchgate.net/publication/356615873
- Great Minds Do Not Think Alike: researchgate.net/publication/359114933
- Tell Us What You Really Think: researchgate.net/publication/370193206