Skip to contents

ruminate

The goal of ruminate is to facilitate exploration of pharmacometrics data. This is done by creating a Shiny interface to different tools for data transformation (dplyr and tidyr), plotting (ggplot2), and noncompartmental analysis (PKNCA). These results can be reported in Excel, Word or PowerPoint. The state of the app can be saved and loaded at a later date. When saved, a script is generated to reproduce the different actions in the Shiny interface.

Deployed example in the cloud

You can point your web browser to runrumiate.ubiquity.tools and try the development version of ruminate. Keep in mind this is a public server. No data is tracked or kept, but it is open to the public. So be careful about uploading confidential data. Also know this is a single instance and if many people are using it at once it can be slow.

Local installation

From CRAN

You can install the released version of ruminate from CRAN with:

# General dependencies:
install.packages("clipr")
install.packages("gridExtra")
install.packages("prompter")
install.packages("readxl")
install.packages("shinydashboard")
install.packages("ubiquity")

# Dependencies for the MB and CTS modules
install.packages("nlmixr2lib")
install.packages("nonmem2rx")
install.packages("rxode2")
install.packages("rxode2et")
install.packages("nlmixr2")

# Actual package
install.packages("ruminate")

Development version

You can install the development version from GitHub with the following:

# Installing devtools if it's not already installed
if(system.file(package="devtools") == ""){
  install.packages("devtools") 
}

# General dependencies:
install.packages("clipr")
install.packages("gridExtra")
install.packages("prompter")
install.packages("readxl")
install.packages("shinydashboard")
install.packages("ubiquity")

# Dependencies for the MB and CTS modules
install.packages("nlmixr2lib")
install.packages("nonmem2rx")
install.packages("rxode2")

devtools::install_github("john-harrold/onbrand",  dependencies=TRUE)
devtools::install_github("john-harrold/formods",  dependencies=TRUE)
devtools::install_github("john-harrold/ruminate")

Note that because ruminate depends on formods and onbrand you will need to first install the development versions of onbrand and formods.

Running ruminate

In RStudio

This will run the default/stable version of ruminate:

As new modules are developed they can be found in the development app. This command will run the ruminate with any development modules enabled:

library(ruminate)
library(shiny)
runApp(system.file(package="ruminate","templates", "ruminate_devel.R"))

If you are running this you should probably be using the development version off of github to make sure you have the latest version.

In a Docker container

This kind of assumes a basic familiarity with Docker. Keep in mind that these containers are built for ‘x86’ images. Essentially Intel/AMD cpus. This means that if you have a Mac with Apple Silicon (M1, M2,etc.) it may not work or the performance may not be that good.

Pull a container from Docker

This will pull container from docker and run it locally

docker pull johnmharrold/ruminate:latest
docker run -p 3838:3838   johnmharrold/ruminate:latest &

If that worked you can put the following into a web browser and then use ruminate there:

http://127.0.0.1:3838/

Create your own container

If you want to create your own container you can do the following:

docker build -t ruminate --no-cache  https://raw.githubusercontent.com/john-harrold/ruminate/main/inst/docker/local_container

This will run the Docker image locally:

docker run --name ruminate --rm -p 3838:3838 ruminate

If everything worked you can then point your web browser to:

http://127.0.0.1:3838/

This will push the image to DockerHub:

docker tag ruminate <username>/ruminate:latest
docker push <username>/ruminate:latest

You can use this to pull the image from DockerHub:

docker pull <username>/ruminate:latest

Deployment and Customization

If you want to deploy and customize ruminate, please see the Deployment vignette.