Some recent publications using Stan


#1

Here are three recent publications from my lab that have used Stan. Code and data on Github for each.

Models are dynamic hierarchical learning models and an IRT analysis.

@article{AplinSheldonMcelreath2017,
URL = {http://www.pnas.org/content/early/2017/07/18/1621067114.abstract},
year = {2017},
author = {Lucy M. Aplin and Ben C. Sheldon and Richard McElreath},
title = {Conformity does not perpetuate suboptimal traditions in a wild population of songbirds},
journal = {PNAS},
Supplemental = {https://github.com/rmcelreath/wythamewa}
}

@article{BunceMcElreath2017,
URL = {https://doi.org/10.1007/s12110-017-9297-8},
year = {2017},
author = {John Bunce and Richard McElreath},
title = {Inter-ethnic Interaction, Strategic Bargaining Power, and the Dynamics of Cultural Norms: A Field Study in an {A}mazonian Population},
journal = {Human Nature}
}

@article{Barrett2017,
doi = {10.1098/rspb.2017.0358},
URL = {https://doi.org/10.1098/rspb.2017.0358},
year = {2017},
month = {jun},
publisher = {The Royal Society},
volume = {284},
number = {1856},
pages = {20170358},
author = {Brendan J. Barrett and Richard L. McElreath and Susan E. Perry},
title = {Pay-off-biased social learning underlies the diffusion of novel extractive foraging traditions in a wild primate},
journal = {Proceedings of the Royal Society B: Biological Sciences}
}


#2

Cool—I love papers wirtten for the queen. And this one’s not even in the statistics sections, which is even cooler!

We should really have an award for awesome-Stan-paper-title-of-the-month!


#3

I just realized that may have sounded sarcastic, which wasn’t at all my intent. These really do sound cool. I wish I had time to dive into more of these papers. I love cognitive science and cognitive anthropology (I did my Ph.D. in cog sci and one of my fave classes as an undergrad was a seminar in cognitive anthropology).

Mitzi and I just spent a weekend with Jeff Benjamin, an old pottery studio mate of Mitzi’s who’s now working on a Ph.D. Industrial Archaeology and Anthropology, which I didn’t even know was a thing. He’s doing really neat work on recreating early industrial soundscapes.


#4

Here’s another that just came out:

@article{KosterMcElreath2017MNAB,
URL = {https://link.springer.com/article/10.1007/s00265-017-2363-8},
year = {2017},
author = {Jeremy Koster and Richard McElreath},
title = {Multinomial analysis of behavior: statistical methods},
journal = {Behavioral Ecology and Sociobiology},
volume = {71},
number = {138}
}

Just hierarchical multinomial regression. Nothing too exciting.

The Royal Society name is indeed anachronistic! Having a scientific society named after an illegitimate, traditional institution of coercive hereditary control seems unnecessary. Here in DE, they renamed the old Kaiser Wilhelm Society (Gesellschaft) to the Max Planck Society (Gesellschaft). But I remember reading that was the result of outside pressure following the aftermath of WW2, as the name Kaiser Wilhelm had become tainted by its association with Nazi research.


#5

Continuing completely off topic.

Working in a subfield named after an 18th century cleric, I try not to throw stones.

Thanks for the background—I had always wondered where all those Max Planck institutes came from. I knew they existed in Germany and the Netherlands, but had no idea they were also in Rome and other places (just looked it up).


#6

Hi Stan users,
We put out a new paper using Stan to predict probability of movement of invasive wild pigs based on anthropogenic covariates. It’s certainly not the most sophisticated use of the program, but it is one of the first uses of Stan in Ecology journals:
Tabak, M. A., A. J. Piaggio, R. S. Miller, R. A. Sweitzer, and H. B. Ernest. 2017. Anthropogenic factors predict movement of an invasive species. Ecosphere 8(6):. 10.1002/ecs2.1844
link


#7

Nice to see more Stan in Ecology! We’re also continuing to use Stan for many projects in my group.

Here is a recent OA example with code provided in supplements to fit a SEM and isotopic mixing model.
@article {Tanentzape1601765,
URL = {http://advances.sciencemag.org/content/3/3/e1601765},
doi = {10.1126/sciadv.1601765},
author = {Tanentzap, Andrew J. and Kielstra, Brian W. and Wilkinson, Grace M. and Berggren, Martin and Craig, Nicola and del Giorgio, Paul A. and Grey, Jonathan and Gunn, John M. and Jones, Stuart E. and Karlsson, Jan and Solomon, Christopher T. and Pace, Michael L.},
title = {Terrestrial support of lake food webs: Synthesis reveals controls over cross-ecosystem resource use},
volume = {3},
number = {3},
pages = {e1601765},
year = {2017},
journal = {Science Advances}
}

Here’s a mechanistic model to predict flowering events in plants, again, with code online:
@article {NPH:NPH13817,
url = {http://dx.doi.org/10.1111/nph.13817},
doi = {10.1111/nph.13817},
author = {Monks, Adrian and Monks, Joanne M. and Tanentzap, Andrew J.},
title = {Resource limitation underlying multiple masting models makes mast seeding sensitive to future climate change},
journal = {New Phytologist},
volume = {210},
number = {2},
pages = {419–430},
year = {2016}
}


#8

And an example of using Stan for generalised mixed models that account for statistical non-independence among data points (species in this case), again with code:

@article {NPH:NPH14167,
url = {http://dx.doi.org/10.1111/nph.14167},
doi = {10.1111/nph.14167},
author = {Tanentzap, Andrew J. and Lee, William G.},
title = {Evolutionary conservatism explains increasing relatedness of plant communities along a flooding gradient},
journal = {New Phytologist},
volume = {213},
number = {2},
pages = {634–644},
year = {2017}
}


#9

That’s great! I hadn’t seen the paper on the isotopic mixing model, but I’ll send this to some colleagues that will be interested.


#10

I’m hoping to be at ISEC in 2018. We didn’t get our act together to propose a tutorial session, but I plan to organize a Stan course in the Edinburgh area between ISBA [24–29 July, 2018, Edinburgh] and ISEC [2–6 July 2018, St. Andrews]. I did my Ph.D. in Edinburgh and go back any chance I can get.

If I do a course, it’ll involve at least a full day of ecology models (mark-recapture, occupation, movement HMMs, etc.—all the stuff I saw being used at ISEC 2016). We have lots of tutorials on ecology models, incuding Hiroki Ito’s translation of Kéry and Schaub, and I wrote a case study replicating Dorazio and Royle’s occupancy model results.


#11

Another paper using Stan (and JAGS), just published in CPT: Pharmacometrics and Systems Pharmacology.

Vandemeulebroecke, Bornkamp, Krahnke, Mielke, Monsch, Quarg: A Longitudinal Item Response Theory Model to Characterize Cognition Over Time in Elderly Subjects. CPT:PSP 2017. (DOI: 10.1002/psp4.12219)

Link: http://onlinelibrary.wiley.com/doi/10.1002/psp4.12219/full


#12

Neat topic. Feel free to start new publication-specific topics and discuss them. The intention was to have a recent publication and publicity topic, not a single thread.


#13

Cool to hear that, you might want to get in touch with some of the Quantitative Ecology or Macroecology folks, I heard that they are planning some annual meeting / workshop around the ISEC meeting in St Andrews. There might be some interest to hear about Stan, HMC and / or developing and maintaining a large computational project …


#14

Thanks for the pointers—I don’t really want to join a mailing list any suggestion of who to contact specifically? I didn’t see any way to scan their existing lists and I don’t want to join another email list.


#15

I would try Natalie Cooper for the macroecology, or Nick Golding for the computational ecology. Both interest groups have their annual meeting after the ISEC in St Andrews.


#16

Thanks. Will do.