Publications by Gavin L. Simpson

Prediction intervals for GLMs part II

01.05.2017

One of my more popular answers on StackOverflow concerns the issue of prediction intervals for a generalized linear model (GLM). Comments, even on StackOverflow, aren’t a good place for a discussion so I thought I’d post something hereon my blog that went into a bit more detail as to why, for some common types of GLMs, prediction intervals ar...

7932 sym R (2635 sym/14 pcs) 4 img 2 tbl

Prediction intervals for GLMs part II

01.05.2017

One of my more popular answers on StackOverflow concerns the issue of prediction intervals for a generalized linear model (GLM). Comments, even on StackOverflow, aren’t a good place for a discussion so I thought I’d post something hereon my blog that went into a bit more detail as to why, for some common types of GLMs, prediction intervals ar...

7932 sym R (2635 sym/14 pcs) 4 img 2 tbl

Fitting count and zero-inflated count GLMMs with mgcv

04.05.2017

A couple of days ago, Mollie Brooks and coauthors posted a preprint on BioRχiv illustrating the use of the glmmTMB R package for fitting zero-inflated GLMMs (Brooks et al., 2017). In the paper, glmmTMB is compared with several other GLMM-fitting packages. mgcv has recently gained the ability to fit a wider range of families beyond the exponentia...

15111 sym R (6451 sym/21 pcs) 10 img

Fitting count and zero-inflated count GLMMs with mgcv

04.05.2017

A couple of days ago, Molly Brooks and coauthors posted a preprint on BioRχiv illustrating the use of the glmmTMB R package for fitting zero-inflated GLMMs (Brooks et al., 2017). In the paper, glmmTMB is compared with several other GLMM-fitting packages. mgcv has recently gained the ability to fit a wider range of families beyond the exponential...

15208 sym R (6451 sym/21 pcs) 10 img

Comparing smooths in factor-smooth interactions I

10.10.2017

One of the really appealing features of the mgcv package for fitting GAMs is the functionality it exposes for fitting quite complex models, models that lie well beyond what many of us may have learned about what GAMs can do. One of those features that I use a lot is the ability to model the smooth effects of some covariate (x) in the different le...

11384 sym R (4285 sym/17 pcs) 6 img

Comparing smooths in factor-smooth interactions I

10.10.2017

One of the really appealing features of the mgcv package for fitting GAMs is the functionality it exposes for fitting quite complex models, models that lie well beyond what many of us may have learned about what GAMs can do. One of those features that I use a lot is the ability to model the smooth effects of some covariate (x) in the different le...

11369 sym R (4285 sym/17 pcs) 6 img

First steps with MRF smooths

19.10.2017

One of the specialist smoother types in the mgcv package is the Markov Random Field (MRF) smooth. This smoother essentially allows you to model spatial data with an intrinsic Gaussian Markov random field (GMRF). GRMFs are often used for spatial data measured over discrete spatial regions. MRFs are quite flexible as you can think about them as rep...

8406 sym R (4425 sym/17 pcs) 8 img

First steps with MRF smooths

19.10.2017

One of the specialist smoother types in the mgcv package is the Markov Random Field (MFR) smooth. This smoother essentially allows you to model spatial data with an intrinsic Gaussian Markov random field (GMRF). GRMFs are often used for spatial data measured over discrete spatial regions. MRFs are quite flexible as you can think about them as rep...

8405 sym R (4425 sym/17 pcs) 8 img

Comparing smooths in factor-smooth interactions II

14.12.2017

In a previous post I looked at an approach for computing the differences between smooths estimated as part of a factor-smooth interaction using s()’s by argument. When a common-or-garden factor variable is passed to by, gam() estimates a separate smooth for each level of the by factor. Using the (Xp) matrix approach, we previously saw that we c...

9438 sym R (3056 sym/12 pcs) 2 img

Fitting GAMs with brms: part 1

21.04.2018

Regular readers will know that I have a somewhat unhealthy relationship with GAMs and the mgcv package. I use these models all the time in my research but recently we’ve been hitting the limits of the range of models that mgcv can fit. So I’ve been looking into alternative ways to fit the GAMs I want to fit but which can handle the kinds of d...

9622 sym R (2853 sym/20 pcs) 10 img