Johannes Brachem
LightNews LightNews
jobrachem.bsky.social
Johannes Brachem
@jobrachem.bsky.social
16 followers 37 following 6 posts
Statistics PhD student, University of Göttingen. Working on Bayesian semiparametric distributional regression.
Posts Replies Media Videos
Reposted by Johannes Brachem
olemole.bsky.social
My paper is out! 🎉 I explore hidden semi-Markov models with covariate-dependent state dwell-time distributions — because sometimes Markov just isn’t enough.
Case study: Arctic muskox movement! 🦬📊
🔗 www.sciencedirect.com/science/arti...

#stats #TimeSeries #HSMM #StatisticalEcology #rstats
Hidden semi-Markov models with inhomogeneous state dwell-time distributions
The well-established methodology for the estimation of hidden semi-Markov models (HSMMs) as hidden Markov models (HMMs) with extended state spaces is …
www.sciencedirect.com
March 19, 2025 at 4:35 PM
Reposted by Johannes Brachem
jvgemert.bsky.social
There's a paper about that :).

The importance of stupidity in scientific research
journals.biologists.com/jcs/article/...
The importance of stupidity in scientific research
I recently saw an old friend for the first time in many years. We had been Ph.D. students at the same time, both studying science, although in different areas. She later dropped out of graduate school...
journals.biologists.com
March 8, 2025 at 7:52 PM
Reposted by Johannes Brachem
sanjaysrivastava.com
Whole journals would go belly up
Don’t write an article with the following structure: * introduction that motivates a causal research question * analysis framed in associational terms * discussion that interprets the findings causally * limitation section that says that findings should not be interpreted causally You can’t have a causal cake and correlate it too.
February 17, 2025 at 3:32 PM
Reposted by Johannes Brachem
andrew.heiss.phd
The viridis palettes are great perceptually uniform color maps for #dataviz, but sometimes at the extremes they're too pale/dark. With ggplot and #rstats, though, you can truncate the palette to avoid the paler colors. I wrote a little guide for how datavizf24.classes.andrewheiss.com/news/2024-12...
When you use scale_fill_viridis_d() or scale_color_viridis_d(), ggplot chops the viridis color map into evenly spaced colors along the range of the gradient. For instance, if you have a plot with 4 categories, you’ll end up with these colors: Figure showing 8 viridis palettes chopped at 0%, 33%, 66%, and 100% R code for creating a bar chart four four continents, with one of the bars filled with a really really pale yellow from the viridis magma palette Figure showing the 8 viridis palettes, but starting at 10% and ending at 80%, resulting in colors at at 10%, 33%, 57%, and 80% R code for creating a bar chart four four continents, with the bars filled with colors from a truncated viridis magma palette because of scale_fill_viridis_d(option = "magma", begin = 0.1, end = 0.8)
December 4, 2024 at 2:43 PM
jobrachem.bsky.social
Excited to finally share this work: We used transformation models to develop location-scale regression for arbitrary conditional distributions! Check out the preprint: arxiv.org/abs/2404.07440 #stats #bayesian
Bayesian Penalized Transformation Models: Structured Additive...
Penalized transformation models (PTMs) are a novel form of location-scale regression. In PTMs, the shape of the response's conditional distribution is estimated directly from the data, and...
arxiv.org
April 12, 2024 at 4:45 PM
jobrachem.bsky.social
If any #rstats folks read this: do you have experience with submitting .tex files generated by Quarto to the ArXiV? Will it just work? If not, how much will I suffer to make it work?
March 26, 2024 at 3:18 AM
Reposted by Johannes Brachem
richarddriley.bsky.social
“Those who ignore Statistics are condemned to reinvent it”
- Bradley Efron.

More relevant now than ever
#AI #machinelearning #ML #DataScience take note
January 31, 2024 at 11:39 AM