Ondřej Mottl | Lightnews

Irena Simova @isimova.bsky.social · 23h

📢 Postdoc opening in avian biogeography & macroecology (Olomouc, CZ)
Vladimír Remeš (excellent scientist and mentor) is hiring a postdoc to study bird diversity across biomes, linking species pools, traits, climate & history using big data, fieldwork, and modelling.
🔗 pracuj.upol.cz/nc/zprava/cl...

Ondrej Mottl 🌿💻📈🌍⏳ @ondrejmottl.bsky.social · 4d

That was the next thing I have done. My application for accessing HPC has just been accepted😉 In CZ, we have #MetaCentrum - a cluster of more than 35000 (!) CPUs...that should do the trick

Ondrej Mottl 🌿💻📈🌍⏳ @ondrejmottl.bsky.social · 4d

I am having issue fitting some large #JSDM with #RStats due to the fact my PC does not enough #RAM (2*32GB DDR5). I thought I will just buy 2 more (have the slots for it)...check the prices... nevermind...

Ondrej Mottl 🌿💻📈🌍⏳ @ondrejmottl.bsky.social · 4d

I have just finished my course on reproducibility in science (#CUNI-SPROuT) but there are some cool ideas to try in the next run👇

Tiffany Timbers @tiffanytimbers.bsky.social · 5d

"Ten simple rules for teaching data science": arxiv.org/abs/2602.02874

A new preprint by @minecr.bsky.social and myself. We'd love any feedback!

Ten simple rules for teaching data science

Teaching data science presents unique challenges and opportunities that cannot be fully addressed by simply borrowing pedagogical strategies from its parent disciplines of statistics and computer scie...

arxiv.org

Ondrej Mottl 🌿💻📈🌍⏳ @ondrejmottl.bsky.social · 5d

Have you tried radian?

Ondrej Mottl 🌿💻📈🌍⏳ @ondrejmottl.bsky.social · 5d

Is it a good one? I was checking it out earlier and was to sure if to dive in (as it seems to be still in alpha/beta).

Tiffany Timbers @tiffanytimbers.bsky.social · 5d

"Ten simple rules for teaching data science": arxiv.org/abs/2602.02874

A new preprint by @minecr.bsky.social and myself. We'd love any feedback!

Ten simple rules for teaching data science

Teaching data science presents unique challenges and opportunities that cannot be fully addressed by simply borrowing pedagogical strategies from its parent disciplines of statistics and computer scie...

arxiv.org

Ondrej Mottl 🌿💻📈🌍⏳ @ondrejmottl.bsky.social · 9d

The more I do teach #RStats, the more I think it is a blessing and a course at the same time.

Brian Helsel @bchelsel.bsky.social · 9d

One of my favorite parts of R is that there are many paths to the same result. A simple task can be written with a for loop, the apply family, purrr, vectorized base R, or even Rcpp. The real skill isn’t memorizing syntax—it’s knowing when to choose each option. #RStats

An image of R code that shows different ways to the mean of a squared vector. The code uses a for loop, the apply family, purrr package, vectorized base R and a C++ function.

Brian Helsel @bchelsel.bsky.social · 9d

One of my favorite parts of R is that there are many paths to the same result. A simple task can be written with a for loop, the apply family, purrr, vectorized base R, or even Rcpp. The real skill isn’t memorizing syntax—it’s knowing when to choose each option. #RStats

Ondrej Mottl 🌿💻📈🌍⏳ @ondrejmottl.bsky.social · 9d

I am always amazed by the amount of useful tools that R can provide. Truly a multipurpose programming language!
#RStats

Andrew Barnas @andrewbarnas.bsky.social · 10d

Took me way too long to realize that its super easy to just merge pdfs in R rather than fussing with Adobe #rstats

library(pdftools)
pdf_combine(
input = c("file1.pdf", "file2.pdf"),
output = "merged.pdf"
)

Andrew Barnas @andrewbarnas.bsky.social · 10d

Took me way too long to realize that its super easy to just merge pdfs in R rather than fussing with Adobe #rstats

library(pdftools)
pdf_combine(
input = c("file1.pdf", "file2.pdf"),
output = "merged.pdf"
)

Ondrej Mottl 🌿💻📈🌍⏳ @ondrejmottl.bsky.social · 15d

This is such a cool idea! I am sorting custom functions into folders but this allows me to do it in env as well!
👏👏👏

John Coene @john-coene.com · 15d

I genuinely had no idea #rstats technically has a module system.

my_module <- new.env()
source("helpers.R", local = my_module)

my_module$add(1, 2)

John Coene @john-coene.com · 15d

I genuinely had no idea #rstats technically has a module system.

my_module <- new.env()
source("helpers.R", local = my_module)

my_module$add(1, 2)

Sivuyile Nzimeni @sivuyilenzimeni.com · 17d

Me, explaining my R script with Base R, data.table and tidyverse code. #rstats

Ondrej Mottl 🌿💻📈🌍⏳ @ondrejmottl.bsky.social · 18d

This is huge!

Bert van der Veen @vdveenb.bsky.social · 18d

#gllvm can now handle multiple response types at the same time (different data types for different species; presence-absence and counts, or counts and overdispersed counts, counts and biomass, ..., for example), thanks to a massive coding exercise by Jenni Niku!

Bert van der Veen @vdveenb.bsky.social · 18d

#gllvm can now handle multiple response types at the same time (different data types for different species; presence-absence and counts, or counts and overdispersed counts, counts and biomass, ..., for example), thanks to a massive coding exercise by Jenni Niku!

Ondrej Mottl 🌿💻📈🌍⏳ @ondrejmottl.bsky.social · 19d

🤯🤯🤯

Nicola Rennie @nrennie.bsky.social · 19d

But sometimes that's not an option. For example, if the plot was created through a function in another package that you can't easily edit. Instead, we can update values stored in the plot object. (Another reason to save your ggplot2 charts as variables instead of just printing the output!)

R code to edit the points layer g$layers[[1]]$aes_params$colour <- "red"

Nicola Rennie @nrennie.bsky.social · 19d

But sometimes that's not an option. For example, if the plot was created through a function in another package that you can't easily edit. Instead, we can update values stored in the plot object. (Another reason to save your ggplot2 charts as variables instead of just printing the output!)

Ondrej Mottl 🌿💻📈🌍⏳ @ondrejmottl.bsky.social · 26d

?
```{r}
purrr::map2(.x=list(data1, data2), .y=c("table 1", "table 2"), .f= ~ tinytable::tt(.x, cap= .y))
```
I would assume that this will make 2 tables with headings

Ondrej Mottl 🌿💻📈🌍⏳ @ondrejmottl.bsky.social · 26d

Maybe it is an overkill. What about just purrr::map into some tinytable::tt()? There is argument 'cap' for header

Ondrej Mottl 🌿💻📈🌍⏳ @ondrejmottl.bsky.social · 26d

Metaprogramming? Make a function (purrr::map) to produce the r chunk directly into quarto doc
stackoverflow.com/a/69538439

Create RMarkdown headers and code chunks in purrr

The following .Rmd is what I think should produce what I'm looking for: --- title: "Untitled" date: "10/9/2021" output: html_document --- ```{r setup, include=FALSE} knitr::

stackoverflow.com

Ondrej Mottl 🌿💻📈🌍⏳ @ondrejmottl.bsky.social · 27d

Thanks for the note! The soil data is only available for contemporary data. Nevertheless, we thought that it is still worth incorporating.

Ondrej Mottl 🌿💻📈🌍⏳ @ondrejmottl.bsky.social · 28d

My experience is that the transition from ECR to leadership role is scary. The sudden freedom and lack of guidance can be overwhelming. What helped me is to surround yourself with similar-minded people and continue doing collaborative work on each other's projects.

Ondrej Mottl 🌿💻📈🌍⏳ @ondrejmottl.bsky.social · 28d

I am interested in how that might have happened. You store stuff only locally? No remote in your git setup?

Ondrej Mottl 🌿💻📈🌍⏳ @ondrejmottl.bsky.social · 28d

10/🧵 Join us!

🔗 Explore the project: bit.ly/VegVault
🐛 Report bugs, suggest features, or let’s collaborate!

Open for feedback & ideas! 💡

a penguin is wrapped in a blanket with the words let 's work together above it

ALT: a penguin is wrapped in a blanket with the words let 's work together above it

media.tenor.com

Ondrej Mottl 🌿💻📈🌍⏳ @ondrejmottl.bsky.social · 28d

9/🧵 Huge thanks to my amazing co-authors 🙌
🤝 Franka Gaiser
🤝 Irena Šímová @isimova.bsky.social
🤝 Suzette G. A. Flantua @suzetteflantua.bsky.social
#DreamTeam

Teamwork GIF

ALT: Teamwork GIF

media.tenor.com

Ondrej Mottl 🌿💻📈🌍⏳ @ondrejmottl.bsky.social · 28d

8/🧵 Meet {vaultkeepr} 🔑🧙‍♂️
Your R interface to VegVault:
✅ no SQL
✅ filter by space/time/taxa/traits
✅ extract data and the citations you must include

#Rstats

🔗 ondrejmottl.github.io/VegVault/web...

Ondrej Mottl 🌿💻📈🌍⏳ @ondrejmottl.bsky.social · 28d

7/🧵 Example vibes ✨
From tracking Picea across North America since the LGM to trait-based reconstructions (e.g. plant height) — the website includes worked examples you can copy/paste & adapt.

🔗 ondrejmottl.github.io/VegVault/web...

$Visualisation of example data usage of VegVault using {vaultkeepr} package. See Usage Notes section for details about the aim of each data compilation. (A) The individual points represent vegetation data (Samples for fossil pollen or contemporary vegetation plots) within each age time slice. (B) Each point represents one contemporary vegetation plot. The colour represents the Mean Annual Temperature (MAT) associated with that plot. Note that several vegetation plots can have the same value due to sharing the same closest gridpoint (see Fig. 2). (C) The left figure represents the temporal length of fossil pollen Datasets, which are available for Central and South America between 6 and 12 ka BP (thousand years before present). The middle figure represents the distribution of various genera present in this data assembly. The right figure represents the distribution of average Plant height Trait per each genus present in the data assembly. The code to reproduce these figures can be accessed from the GitHub repository titled VegVault (https://github.com/OndrejMottl/VegVault/tree/v1.0.0-website_rewiev), which can be accessed as https://doi.org/10.5281/zenodo.16987345.$

Ondrej Mottl 🌿💻📈🌍⏳ @ondrejmottl.bsky.social · 28d

6/🧵 Abiotic linking 🌦️🪨

VegVault connects records to a spatio-temporal grid of climate/soil “gridpoints” — so you can pull drivers without painful raster wrangling.

🔗 ondrejmottl.github.io/VegVault/web...

Example of the resolution of the gridpoint Datasets extracting abiotic information. (A) Map of Europe. The red line indicates country borders with the Czech Republic as the focus of the map. (B) Distribution of the individual gridpoints (grey squares) within the Czech Republic (red). (C) Linking vegetation plots and gridpoints. The green circles represent the contemporary vegetation plots. The grey lines connect the vegetation points to their closest gridpoint. Note that the abiotic data is provided to each vegetation plot from their closest gridpoint (i.e., the vegetation plots connected to the same gridpoint will have the same abiotic data). Gridpoints without vegetation plots in their spatio-temporal vicinity (here contemporary data) were excluded from the VegVault to reduce the database size.

Ondrej Mottl 🌿💻📈🌍⏳ @ondrejmottl.bsky.social · 28d

5/🧵 Scale check 🤯🌍
480k datasets, >13M samples, >100k taxa + millions of trait values — spanning space and millennia. This is what “macro-ecology across time” should feel like 😎

🔗 ondrejmottl.github.io/VegVault/web...

#Global #Palaeobiology

Spatio-temporal distribution of vegetation Samples (for both vegetation plots and fossil pollen archives). Note that this is overview of current state of VegVault (v1.0.0.). A: Spatial distribution with each point representing one plot. B: Temporal distribution of individual Samples

Add to Home Screen

Light up
your news

Add to Home Screen

Light upyour news

Sign in to Lightnews

Sign up to start reading

Connect Bluesky

Connect with Bluesky

Light up
your news