A bunch of R packages
An uscimage R package with rmarkdown templates
The Happy Scientist Semminars
The Writing R packages Workshop
We have a list of github issue templates that can be used as checklists for starting projects.
We have AWS, now what...
Bioghost 2: On the process of acquiring new hard drives to make both machines bioghost 1/2 to use the same filesys.
Upcomming Happy Scientist Semminars: Shiny, Docker (?)
The Intro to HPC with R
Sort of alliance with HPCC for education (possible will be doing a workshop on R by the end of the summer)
A list of R packages (more here):
(*) New since the EAC
Working version of the MCMC/MLE estimators
Simulations using PANTHER data showed the MCMC has good coverage and is performing as expected.
Working and highly tested R package (but the posterior probabilities)
Looking at the last point, a corrected and throughly tested algorithm for computing posterior probabilities (the jmorr tests)
We have included a new parameter in the model, $\eta$ publication bias probabilities.
New version of the R package allows more flexibility in model specification. The likelihood function is written based on formulas:
```r
dat ~ mu
dat ~ mu + psi
dat ~ mu + psi(0, 1)
dat ~ mu + psi + eta + Pi ```
We are working on re-running the algorithm on both simulated annotations and manually curated data.
A couple of examples ```r
library(aphylo)
set.seed(3) dat <- raphylo(40, eta = c(.7, .9))
plot(dat)
ans <- aphylo_mcmc( dat ~ psi + eta, control = list(nbatch = 5e4, nsteps=4, thin=10), priors = function(p) c( # Beta priors for all parameters dbeta(p[c("mu0", "mu1", "psi0", "psi1")], 2, 18), dbeta(p[c("eta0", "eta1")], 18, 2) ) )
ans plot(ans) # Looking at the surface
pred <- predict(ans) prediction_score(ans)
plot(prediction_score(ans)) ```
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.