Russell Dinnage

autograd functional?

9

I am interested in some of the functionality of torch.autograd.functional (https://github.com/pytorch/pytorch/blob/master/torch/autograd/functional.py) in Pytorch for R (especially `jvp()` and `jacobian()`). Is there plans for adding this functionality to R `torch`? I...

This is a system requirement for an upcoming package I'm writing, so I thought I'd contribute an entry.

Cannot save any torch tensor or `nn_module`

4

Ever since I updated `torch` to the latest version, I have not been able to save any `torch` object using the default serialization method. I have instead had to fall...

Support `emmeans`

1

Support the `emmeans` package for estimating marginal means from `communityPGLMM` objects. There is a vignette for developers here: https://cran.r-project.org/web/packages/emmeans/vignettes/xtending.html On first glance this shouldn't be too difficult?

Migrate testing from travis to github actions?

1

@daijiang I notice that testing is currently handled by travis ci (and the current build seems to be failing?). I think most packages these days use github actions instead. I...

Memory efficient version of phylogenetic diversity metrics

5

A PSV/PSE/PSC/etc version that can handle really big phylogenies. In the past I have tried to calculate PSV on a phylogeny with several hundred thousand tips, but R will give...

We removed `ppmlasso` when it got kicked from CRAN (#197). It looks like it is back up and updated now (https://cran.r-project.org/web/packages/ppmlasso/index.html), so should probably look into getting it back into...

Add options to `enmtools.ppmlasso` to do more of the advanced features of `ppmlasso`, such as point interaction models, which allows occurrence points to 'interact', e.g. attract or repulse each other...

formula recasting doesn't allow suppressing intercept

1

I noticed that if a user passes in a formula like this `presence ~ 0 + bio1` or `presence ~ bio1 -1`, the formula recasting (`f

Add cox process SDM using r-inla

4

I discovered a cool new SDM method. Using a cox process model in the r-inla package (https://www.math.ntnu.no/inla/r-inla.org/tutorials/spde/spde-tutorial.pdf; see Chapter 4). r-inla uses integrated nested laplace approximation to fit a Bayesian...