Ben Augustine
@bencaugustine.bsky.social
Ecological Statistics. I came for the charismatic megafauna and stayed for the problematic megamodels. Skeptic. github.com/benaug
Passing those posterior predictive checks
a man wearing 3d glasses and a hard hat says " like a boss "
ALT: a man wearing 3d glasses and a hard hat says " like a boss "
media.tenor.com
November 24, 2025 at 7:19 PM
Passing those posterior predictive checks
Recently came across this Dorazio et al. paper showing that parameter expansion can make probit multispecies occupancy models much more efficient to sample.
besjournals.onlinelibrary.wiley.com/doi/full/10....
I wanted to learn a new trick so I made a nimble version.
github.com/benaug/Probi...
besjournals.onlinelibrary.wiley.com/doi/full/10....
I wanted to learn a new trick so I made a nimble version.
github.com/benaug/Probi...
November 19, 2025 at 6:23 PM
Recently came across this Dorazio et al. paper showing that parameter expansion can make probit multispecies occupancy models much more efficient to sample.
besjournals.onlinelibrary.wiley.com/doi/full/10....
I wanted to learn a new trick so I made a nimble version.
github.com/benaug/Probi...
besjournals.onlinelibrary.wiley.com/doi/full/10....
I wanted to learn a new trick so I made a nimble version.
github.com/benaug/Probi...
Today I learned that people will just remove individuals with large home range sizes from their SCR analyses and not tell anyone
November 19, 2025 at 4:20 PM
Today I learned that people will just remove individuals with large home range sizes from their SCR analyses and not tell anyone
Reposted by Ben Augustine
COME WORK WITH ME!!!
Very exciting open faculty position at MIT, with ecosystem and biodiversity modeling as a focus!
cee.mit.edu/people/share...
Very exciting open faculty position at MIT, with ecosystem and biodiversity modeling as a focus!
cee.mit.edu/people/share...
November 3, 2025 at 7:01 PM
COME WORK WITH ME!!!
Very exciting open faculty position at MIT, with ecosystem and biodiversity modeling as a focus!
cee.mit.edu/people/share...
Very exciting open faculty position at MIT, with ecosystem and biodiversity modeling as a focus!
cee.mit.edu/people/share...
TIL that the slice sampler in jags can't handle unbounded support so jags uses MH that adds/subtracts 1 integer at a time to update latent poisson RVs etc. 🤯 Nimble FTW again.
October 31, 2025 at 7:33 PM
TIL that the slice sampler in jags can't handle unbounded support so jags uses MH that adds/subtracts 1 integer at a time to update latent poisson RVs etc. 🤯 Nimble FTW again.
added a more rigorous explanation of my N-prior data augmentation approach
github.com
October 29, 2025 at 3:03 PM
added a more rigorous explanation of my N-prior data augmentation approach
I was today years old when I noticed that Bayesian p-values for SCR models with data augmentation arbitrarily depend on the level of data augmentation because the simulated test statistic is always greater than or equal to the observed test statistic when z_i=0. Has anyone pointed this out before?
October 27, 2025 at 4:07 PM
I was today years old when I noticed that Bayesian p-values for SCR models with data augmentation arbitrarily depend on the level of data augmentation because the simulated test statistic is always greater than or equal to the observed test statistic when z_i=0. Has anyone pointed this out before?
Lazy Coding Sunday
October 26, 2025 at 5:01 PM
Lazy Coding Sunday
The Royle-Young SCR model never saw much use, in fact I'm not sure if anyone used it outside of the original paper. Maybe bc it was too basic to be widely applicable and making it more flexible required the flexibility of nimble. Anyways, excited about an application that might start its comeback.
October 16, 2025 at 10:21 PM
The Royle-Young SCR model never saw much use, in fact I'm not sure if anyone used it outside of the original paper. Maybe bc it was too basic to be widely applicable and making it more flexible required the flexibility of nimble. Anyways, excited about an application that might start its comeback.
After all the study planning, field work, model development/ refinement, and computation--Behold! the posterior distribution
October 15, 2025 at 3:05 PM
After all the study planning, field work, model development/ refinement, and computation--Behold! the posterior distribution
Reposted by Ben Augustine
Incorporating habitat selection does not account for nonrandom camera deployment in a design-based viewshed density estimator https://www.biorxiv.org/content/10.1101/2025.08.22.671559v1
August 27, 2025 at 7:31 PM
Incorporating habitat selection does not account for nonrandom camera deployment in a design-based viewshed density estimator https://www.biorxiv.org/content/10.1101/2025.08.22.671559v1
All the cool quantitative ecology jobs are postdocs
October 2, 2025 at 1:56 PM
All the cool quantitative ecology jobs are postdocs
Reposted by Ben Augustine
Thinking about all of my US colleagues today that work for USGS, USFWS, US Parks Service, NOAA and the EPA. From front line environmental practitioners to scientists and policy makers, these folks ensure that waters are swimmable, fish are edible, that natural areas are accessible, etc, etc, etc.
October 1, 2025 at 11:24 AM
Thinking about all of my US colleagues today that work for USGS, USFWS, US Parks Service, NOAA and the EPA. From front line environmental practitioners to scientists and policy makers, these folks ensure that waters are swimmable, fish are edible, that natural areas are accessible, etc, etc, etc.
Matern Type III repulsive point processes in NIMBLE. Pretty cool approach and does not require a birth-death process for estimation, so it is fast and mixes well. Can add SCR observation model.
github.com/benaug/Mater...
github.com/benaug/Mater...
September 11, 2025 at 7:58 PM
Matern Type III repulsive point processes in NIMBLE. Pretty cool approach and does not require a birth-death process for estimation, so it is fast and mixes well. Can add SCR observation model.
github.com/benaug/Mater...
github.com/benaug/Mater...
Fun with multimodal posteriors
September 11, 2025 at 3:17 PM
Fun with multimodal posteriors
Can't believe no one thought to make rules about skinning bears in the dorm kitchen 😂
www.cornellsun.com/article/2025...
www.cornellsun.com/article/2025...
Cornell Students Skin Bear in Ganędagǫ: Hall Communal Kitchen, No Charges Filed
Two Cornell students face no charges after skinning a bear in a communal kitchen in the Ganędagǫ: Hall student residence hall on Saturday.
www.cornellsun.com
September 9, 2025 at 2:28 PM
Can't believe no one thought to make rules about skinning bears in the dorm kitchen 😂
www.cornellsun.com/article/2025...
www.cornellsun.com/article/2025...
Got generalized Matern Type III thinning process for point repulsion working using data augmentation. Cool approach, but pretty slow.
August 26, 2025 at 1:09 PM
Got generalized Matern Type III thinning process for point repulsion working using data augmentation. Cool approach, but pretty slow.
Term extended for a full year. Whew!
August 21, 2025 at 7:15 PM
Term extended for a full year. Whew!
Figured out how this data augmentation approach to is related to that presented by Schofield and Barker (2014). Theirs can be specified with bugs code, which is very slick! As coded, mine runs 2x as fast with better ESS/iteration. Cool stuff. More formal details soon.
github.com/benaug/SCR-N...
github.com/benaug/SCR-N...
August 13, 2025 at 4:32 PM
Figured out how this data augmentation approach to is related to that presented by Schofield and Barker (2014). Theirs can be specified with bugs code, which is very slick! As coded, mine runs 2x as fast with better ESS/iteration. Cool stuff. More formal details soon.
github.com/benaug/SCR-N...
github.com/benaug/SCR-N...
Setting up final run for SMRing ground squirrels. Mb for marking process, negative binomial for sighting process, both sex-specific.
August 2, 2025 at 2:30 PM
Setting up final run for SMRing ground squirrels. Mb for marking process, negative binomial for sighting process, both sex-specific.
I have two moods, "everything is cool" and "there is a bug in my code". Everything is cool again.
August 2, 2025 at 1:04 PM
I have two moods, "everything is cool" and "there is a bug in my code". Everything is cool again.
It's Term Renewal Day! Do I still have a job? Who knows! Not my supervisors.
July 25, 2025 at 2:42 PM
It's Term Renewal Day! Do I still have a job? Who knows! Not my supervisors.
Reposted by Ben Augustine
The WildCo Lab @wildco.bsky.social is recruiting 2 postdocs in quantitative ecology to work on mammal population estimation and monitoring from camera trap data. 🐺🐻🦌📷📈
Please share the word or apply to join us at UBC in lovely Vancouver, Canada!
wildlife.sites.olt.ubc.ca/files/2025/0...
Please share the word or apply to join us at UBC in lovely Vancouver, Canada!
wildlife.sites.olt.ubc.ca/files/2025/0...
July 8, 2025 at 11:56 PM
The WildCo Lab @wildco.bsky.social is recruiting 2 postdocs in quantitative ecology to work on mammal population estimation and monitoring from camera trap data. 🐺🐻🦌📷📈
Please share the word or apply to join us at UBC in lovely Vancouver, Canada!
wildlife.sites.olt.ubc.ca/files/2025/0...
Please share the word or apply to join us at UBC in lovely Vancouver, Canada!
wildlife.sites.olt.ubc.ca/files/2025/0...
The overdispersion is strong in the SMR data set
July 12, 2025 at 12:46 PM
The overdispersion is strong in the SMR data set