#easystats makes #rstats easy

RE: https://bsky.app/profile/did:plc:kwjddjickma3syepdsfufyvi/post/3lqskdxzgic2g

A cool feature is that `demean()` also works for nested or cross-classified data structures (when you work with mixed models), i.e. person-mean centering takes your model/data structure into account. Details in `?demean`. #easystats #rstats

RE: https://bsky.app/profile/did:plc:5kx3l44skokbyab6ycny437w/post/3ljzceojun22n

Also interesting: the analytic workflow vignettes: easystats.github.io/easystats/ar... These are short summaries of common workflows when doing data analysis/research... #easystats #rstats

RE: https://bsky.app/profile/did:plc:5kx3l44skokbyab6ycny437w/post/3li2aorawc22f

Articles

We put a lot of work into this update, making @vincentab.bsky.social {marginaleffects} the default backend for estimating marginal means, contrasts and slopes. This was the "missing puzzle part" in our #easystats ecosystem, and it's great! Understand your regression models, made easy! #rstats

RE: https://bsky.app/profile/did:plc:5kx3l44skokbyab6ycny437w/post/3lhhrtikwmc2q

Oh yes, next {modelbased} version will be great! #rstats #easystats

RE: https://bsky.app/profile/did:plc:5kx3l44skokbyab6ycny437w/post/3lgr3qf4wmc2z

The #easystats team rarely meets in person but the best ideas always come out of it to expand and develop our collection of #rstats @rstats packages!

Here you can see the switch from RStudio to Positron by @rempsyc as guided by @strengejacke 😎

Wallpapers are the new hex stickers #rstats #easystats

But still, an impressive amount of news for this release: easystats.github.io/insight/news... #rstats #easystats

RE: https://bsky.app/profile/did:plc:5kx3l44skokbyab6ycny437w/post/3lbucui7rxs2e

Changelog

Would love to hear some opinions on this: which information is useful in the output, which information should just go into the docs? #rstats #easystats

RE: https://bsky.app/profile/did:plc:5kx3l44skokbyab6ycny437w/post/3l7if4jomhv2z

New publication in @psychonomic_soc Behavior Research Methods!

We dive deep into simplifying outlier detection in R using #easystats to follow good practices and make your data analysis more robust and replicable. Check it out! #Rstats #DataAnalysis @rstats

https://doi.org/10.3758/s13428-024-02356-w

@strengejacke Sorry, Daniel! I responded to an old post:

“Is it possible to export a correlation matrix from #rstats to Excel? Yes. With colour based on the correlation coefficient for #dataviz? Yes.
library(rempsyc)
cormatrix_excel(infert, file.name = "mymatrix")
A huge thanks to 
#easystats's 'correlation' package, and Jan Marvin Garbuszus's 'openxlsx2' package, the real powerhouses making this possible!
easystats: @strengejacke, @wiernik, @mattansb and others!”

The #rstats datawizard package (from the #easystats ecosystem) has two very useful functions to deal with duplicates.

* data_duplicated: Extract all duplicates including the first, unlike duplicated() or dplyr::distinct()

* data_unique: by default selects the "best" duplicate

- All regression functions now provide a 95% Ci for the index of effect size, the semi-partial correlation squared (sr2), thanks to an update in
#easystats 's {effectsize} package
- New function to visualize outliers: plot_outliers()
- New function to handle more complex contrast models, nice_lm_contrasts()
- nice_assumptions() & nice_var() now supports list objects

Report version 0.5.6 is now on CRAN!

This version includes improvements to `report_sample()` and `report_effectsize()`, and bumps the minimum needed #rstats version to 3.6.

Oh yeah... and I'm a proud new member of
#easystats
& the new maintainer for report :)

https://easystats.github.io/report/index.html

Thanks for the warm welcome @lakens. This is my #introduction.

I am a PhD Candidate in psychology based in Montreal. My research generally explores ways to impact implicit social cognition and boost prosociality, with tools like meditation and VR, while looking at moderators, like self-control.

I have a passion for #rstats, and authored two such packages: rempsyc (general functions for statistics and plotting), and lavaanExtra (SEM). I am also a very proud and very junior member of #easystats.

The check_outliers() function from
#easystats performance #rstats package now has a dramatically improved output 😍

It now prints: (1) the number of outliers, (2) their row number, (3) the methods and thresholds used, (4) which outliers were flagged for more than 1 variable/method, and (5) their row ID.

Is it possible to export a correlation matrix from #rstats to Excel? Yes. With colour based on the correlation coefficient for #dataviz? Yes.

library(rempsyc)
cormatrix_excel(infert, file.name = "mymatrix")

A huge thanks to
#easystats's 'correlation' package, and Jan Marvin Garbuszus's 'openxlsx2' package, the real powerhouses making this possible!

easystats: @strengejacke, @wiernik, @mattansb and others!

And because we like it so much, here's an example for different types of models, including mixed models. There was a bug for some edge case mixed models, so you might want to get the latest package version of {parameters}, which you can achieve by running following code: `easystats::install_latest(force = TRUE)` (requires package {easystats} from CRAN) #easystats #rstats

The same works for multiple models, too... #easystats #rstats

the #easystats {parameters} package was updated on CRAN. One of the improvements focused on printing tables. You can see the different ways of generating tables here: https://easystats.github.io/parameters/articles/model_parameters_print.html

However, we also improved the HTML-printing, which was a bit behind the textual printing regarding feature completeness, and this gap should be relatively closed now... #rstats