The 3rd chapter of my PhD is out! 🌊
Migratory species strongly affect seabird biomass in seasonal assemblages off northeast Aotearoa/New Zealand - J. Mar. Syst. - doi.org/10.1016/j.jm...
#Seabirds #Oceanography #CommunityEcology #Biomass #GLLVM
Registration is now open for the Physalia online course GLLVMs: Advanced Multivariate Analysis in R (7–10 July) led by
@vdveenb.bsky.social
Join us to learn model-based ordination, JSDMs, and multispecies GL(M)Ms. Hands-on in R!
www.physalia-courses.org/courses-work...
#Ecology #RStats #GLLVM
Quick question for @vdveenb.bsky.social: is there a rough estimate for when the next CRAN update for #gllvm is likely? I want to explore mixed response type models (v.2.0.7), but am currently unable to install the development version at work (CRAN version is v.2.0.5) Thanks very much!
New article on #gllvm by Audun Rugstad.
An approachable guide (equations omitted) on model-based ordination for community ecologists.
Full of practical tips and recommendations, examples, and reoroducible code in the appendix.
doi.org/10.32942/X2K...
As @bbolker.bsky.social indicated, you can do this now in #gllvm. Happy to have a chat if you want to explore that further.
#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!
OK, #gllvm now finally provides an se.fit option in the prediction function, to get simulated confidence intervals.
This was a 5-years old request that should have probably been picked up sooner or later, but better late than never?
Unfortunately, I think I've already reached my conference quotum for next year. Someone should really go and convince the vegetation crowd that model-based ordination is the One True Way #GLLVM
#gllvm is constantly evolving and expanding, so Pekka Korhonen wrote a paper demonstrating new features in the package since Niku et al. 2019.
peerj.com/articles/203...
We've already been working on even more features, so I suppose eventually a 3.0 paper will have to follow... 🙃
In recent time we have greatly expanded the available distributions in #gllvm. Particularly with ZIB, ZNIB, ZINB, negative-binomial (1), and logit and cloglog link functions for the binomials.
We continue to expand the support for VA. But, if all else fails, LA is also available in #gllvm.
There are different methods to estimate GLMMs; the aforementioned packages fit with Laplace approximation (LA). Some other packages use PQL. #gllvm fits models by default with (Gaussian) Variational Approximation (VA).
Just added a zero-inflated binomial option to #gllvm. ZIP, ZINB, ZIBeta were already supported. At some point it would be cool to fully extend this into model-based ordination of zero-inflated components (similar to #glmmTMB and multispecies occupancy).
If I could only find the time..
Pretty cool; Lai (2025, besjournals.onlinelibrary.wiley.com/doi/10.1111/...) advocates for using GLLVMs in phenological studies. They demonstrate with the #glmmTMB package while referring to the #gllvm package for temporally structuring LVs, relaxing the temporal independence assumption in glmmTMB.
Thanks @physaliacourses.bsky.social for again facilitating me in teaching this #GLLVM workshop. The feedback from the participants last year really improved the material. The group this year was so interactive! Looking forward to do this again next year 😁.
For people working with compositional data (such as from eDNA), Wanqi Tang, @sarataskinen.bsky.social et al. compared the performance of a few methods, including of model-based ordination #GLLVM
jdssv.org/index.php/jd...
@bbolker.bsky.social I have a random probit model, and #glmmTMB fits much slower than #gllvm (see below with VA, LA is about half the time of glmmTMB). Any idea what's going on?
The summer school in model-based multivariate analysis is getting closer! The invites have now been sent, time to start preparing the content.. #GLLVM
New version of the gllvm package just hit CRAN! Mostly a lot of bug fixes post Major changes for version 2.0.0. #GLLVM
I am excited to announce that we are advertising a summer school in model-based multivariate analysis for ecologists. It will take place in Hjerkinn, Norway from 21-25 July.
More information can be found at bertv.folk.ntnu.no/index.html and in this 🧵.
#GLLVM
The description for the #GLLVM workshop at #OikosNorway is now on the conference website.
If your student wants to learn how to use the best multivariate analysis methods in community ecology, you should probably send them our way.
I'm giving an invited talk at the conference of the Dutch association for ordination and classification tomorrow. The slides can be found here:
github.com/BertvanderVe...
#GLLVM
