Skip to contents

Installation

pcvr v1.0.0

Source: vignettes/articles/installation.Rmd
installation.Rmd

Installing R

If you do not already have R installed on your local computer then please install R from the comprehensive R archive network (CRAN).

Installing Rstudio

If you do not have Rstudio or some IDE that you like to work with installed then please install Rstudio using the instructions here.

Installing pcvr

Stable Release

A stable version of pcvr can be installed from CRAN using:

Development version

Alternatively the development version can be installed from github using devtools or remotes through R. First install either devtools or remotes from CRAN:

The devtools package makes developing your own R packages more straightforward and lets you easily install packages from places other than CRAN.

With devtools installed we can use install_github to install a more frequently updated version of pcvr from the Danforth center github account. By default this will install package dependencies and will not build the vignettes. See the section in this document for examples of different installation options or the pkgdown site for online docs.

devtools::install_github("danforthcenter/pcvr")

Test installation

Just to check that pcvr is installed and that your R session knows where to find it try running these 3 lines to load the package and generate a plot of some simulated data.

library(pcvr)
priors <- list("A" = c(100, 130), "B" = c(10, 8), "C" = c(0.2, 0.1))
plotPrior(priors, "gompertz")[[1]]

Example plot of default prior distributions to test that the package is working.

Optional Extras

By default install_github will install the dependencies for pcvr, but there are a handful of packages that are not true dependencies but which are very helpful for certain tasks. Currently the brms, cmdstanr and caret packages fall into this category, with brms (Bayesian Regression and Modeling using Stan) and cmdstanr being used for Bayesian growth modeling and caret begin used in partial least squares regression (PLSR). You can install all “Suggested” packages as well by using:

devtools::install_github("danforthcenter/pcvr", dependencies = TRUE)

Alternatively you can just install brms and cmdstanr using instructions below.

brms specifies models in Stan which are then “transpiled” and executed using C++, so you need an interface to make R communicate with Stan/C++, rstan and cmdstanr are good options with cmdstanr generally having better features at this time.

install.packages("cmdstanr", repos = c("https://mc-stan.org/r-packages/", getOption("repos")))
cmdstanr::install_cmdstan()

Test cmdstanr

A simple model can be used to test the cmdstanr installation:

?cmdstanr::cmdstanr_example
cmdstanr::cmdstanr_example()

Test brms

And now we can test the brms installation. Here we use a toy model of the mtcars dataset then a pcvr example.

library(brms)
data(mtcars)
m <- brm(mpg ~ gear + hp, iter = 500, data = mtcars, backend = "cmdstanr")
simdf <- growthSim("linear",
  n = 20, t = 25,
  params = list("A" = c(1.1, 0.95))
)
ss <- growthSS(
  model = "linear", form = y ~ time | id / group,
  sigma = "linear", df = simdf, priors = list("A" = 1)
)
fit_test <- fitGrowth(ss,
  iter = 1000, cores = 2, chains = 2,
  backend = "cmdstanr"
)

pcvr Vignette

The pcvr Vignette shows an example workflow to analyze single and multi value traits from plantCV output. The example data comes from a longitudinal experiment, but most of what is shown would be applicable to other designs as well. The vignette is not automatically built, but can be built on installation:

devtools::install_github("danforthcenter/pcvr", build_vignettes = TRUE)

Now check the available vignettes in your browser with browseVignettes