search

Claudius-Appel / duflor.gui

GUI frontent for `duflor`-package
https://claudius-appel.github.io/duflor.gui/
MIT License
0 stars 0 forks source link
imageanalysis plant-growth r

readme

duflor.gui

R-CMD-check: Dev

The duflor.gui-package contains a shiny-application built for the use of the duflor-package.

Installation - from within RStudio

This package requires the installation of the duflor-package, which implements most of the back-end code for the analysis-pipelines in this application.

Prerequisites

As this application is quite complex, it relies on a certain number of packages to work. The package renv can help with managing a reproducible package environment for a project. As an additional advantage, this also prevents packages in your system-library from being updated when installing this package.

For a proper documentation and information on its use-cases, refer to “https://rstudio.github.io/renv/index.html”.

Setup r-project with renv

To set up a stand-alone R-project with renv, the easiest way is through R-Studio:

  1. Go to File > New Project.
  2. Select New Directory.
  3. Select New Project as project-type[^1].
  4. Give a distinct name to the project, e.g. duflor.gui_installed[^2].
  5. Choose a location for the R-project.
  6. Make sure that the check-box Use renv with this project is checked.[^3]
  7. Make sure the check-box Open in new session is checked.
  8. Create the project.

From now on, this project can be opened from within R-Studio via the project-selection drop-down-menu in the top-right corner of the RStudio-application.

Before continuing, open this project if you did not do so.

Install devtools

First of all, make sure to have followed the steps outlined in “[Setup r-project with renv]”, and that RStudio is currently in your created project.

Installing this package from GitHub requires the installation of the R-package devtools. devtools can be used to compile source-packages to installed binaries.

install.packages("devtools")

When being asked to proceed, make sure that the installation-path is in the R-project you created:

If you created your R-project with renv at

C:/Users/User_Main/Desktop/TempTemporal/test_duflorgui_install

, the path would look something like

C:/Users/User_Main/Desktop/TempTemporal/test_duflorgui_install/renv/library/R-X.Y/....

If this is correct, you can go ahead and answer the prompt with Y. If this is not the case, install devtools via the RStudio-GUI:

  1. Open the packages-panel.
  2. Search for and select devtools.
  3. Under “Install to library”, make sure the path selected is similar to the example outlined above. In other words, make sure you do not install to the library containing the string R/cache/R/renv/sandbox.

Once you have installed devtools, go on to the next step.

Install R-build-toolchain

The underlying package duflor contains C++ code. As it is distributed as source-code, users will require the appropriate R-build-toolchain for their respective operating system. For further information and setup-instructions, refer to “https://r-pkgs.org/setup.html#setup-tools”.

For windows, the R-package installr may be used to check if Rtools is already installed, and whether or not it has been found by R. Additionally, it will notify the user if a new R-update is available (major and minor versions only, ignores patch versions):

install.packages("installr")
# once pkgbuild is installed, run:
installr::install.Rtools(check = T,check_r_update = T,GUI = T)

Installation-steps

Both the duflor- and duflor.gui-packages must be installed from their respective GitHub-repositories.

You can install the development and release versions of duflor.gui like so:

# release:
devtools::install_github("https://github.com/Claudius-Appel/duflor.gui@master")
# development version:
devtools::install_github("https://github.com/Claudius-Appel/duflor.gui@dev")

As with installing devtools, make sure the renv-library is selected when installing the package.

Installing help-vignettes

The application is documented in detail via pkgdown, and its site can be reached from the github-repository.

Most of the articles outlined on the site may also be viewed as vignettes from within RStudio. To do so, you must explicitly declare vignettes to be installed:

devtools::install_github("https://github.com/Claudius-Appel/duflor.gui@master"
  ,build_vignettes = T)
devtools::install_github("https://github.com/Claudius-Appel/duflor.gui@dev"
  ,build_vignettes = T)

Vignettes are rendered documents discussing specific topics within a package. Installing them is recommended, but not required for utilizing the package itself.

Installation - from the command-line

Prerequisites

CLI-Installation-steps

To install the application, follow the steps below:

  1. open the terminal, and move to the location where you want to install the program
  2. create a folder to place the application in, e.g. mkdir duflor_app
  3. move into the folder via cd duflor_app
  4. launch R by executing R
  5. executing 1+1 should now successfully return [1] 2
  6. check if the package renv is installed by executing system.file(package='renv')
    • if it is not, install it by executing install.packages("renv"), then follow instructions for installing it/potential dependencies
  7. execute renv::init()
  8. repeat step 6) for the package devtools
  9. install duflor.gui by executing either:
    • release-version: devtools::install_github("https://github.com/Claudius-Appel/duflor.gui@master")
    • development-version: devtools::install_github("https://github.com/Claudius-Appel/duflor.gui@dev")
    • follow instructions for installing it/potential dependencies
  10. once it is installed, try executing duflor.gui::duflor_gui() to see if the package was installed successfully
    • If it was, the GUI should open in the default-browser, and the application can be used from now on

Installing help-vignettes

The application is documented in detail via pkgdown, and its site can be reached from the github-repository.

Most of the articles outlined on the site may also be viewed as vignettes from within RStudio. To do so, you must explicitly declare vignettes to be installed when performing step 9 in “CLI-Installation-steps”:

devtools::install_github("https://github.com/Claudius-Appel/duflor.gui@master",build_vignettes = T)
devtools::install_github("https://github.com/Claudius-Appel/duflor.gui@dev",build_vignettes = T)

Vignettes are rendered documents discussing specific topics within a package. Installing them is recommended, but not required for utilizing the package itself.

Running the app from the command-line

After installation, the application can be launched as described below:

  1. cd into the folder in which you have installed the application in, e.g. cd duflor_app
  2. launch R via R
  3. launch the app via duflor.gui::duflor_gui()

Running the app from within RStudio

  1. Follow the installation-steps outlined in “Installation - from within RStudio”
  2. Open the project via the drop-down-menu in the top-right corner of the RStudio-application.
  3. Launch the app by executing duflor.gui::duflor_gui() in the console.

Parallelisation requires installation of package, not just loading

In order for the shiny application to be able to run the analysis in parallel, the packages duflor and duflor.gui must be installed. If this package is developed locally, executing devtools::load_all() will not be sufficient. To counteract this, make sure that both packages are actually installed as such, and show up in the listed packages.

[^1]: This might sound confusing, as this is a R-package wrapping a shiny-application. However, those two options are for developing either of those, not for installing them.

[^2]: The name is not really important, but it should be distinct and indicate that this is the installed application

[^3]: If renv is not installed/must be updated, you should be prompted to install it here.