Pathmodelr is an -in development- package for the R language. It implements the Process PLS algorithm and provides helper functions to allow you to perform analysis as efficiently as possible!
The Pathmodelr package is still under development, and as such not yet released through CRAN. To work with the Pathmodelr package (at your own risk!) you can follow the following instructions to set it up. We provide installation instructions for our pre-built package as well as directly from the Github repo using devtools.
First, download the latest release from Github and save it to your disk. Alternatively, it can be downloaded from https://data.mendeley.com/datasets/9x9h7fr4kn
The dependencies will then need to be installed. These dependencies are: 'caret', 'dplyr', 'ggplot2', 'listenv', 'reshape2', 'R6', 'network', 'ggnetwork', 'diagram'
You can install them by running install_packages('package_name') for each package manually, or by using the following command:
install.packages(c('caret', 'dplyr', 'ggplot2', 'listenv', 'reshape2', 'R6', 'network', 'ggnetwork', 'diagram'))
After the dependencies are isntalled you can install the package using the following command:
install.packages("PATH_TO_RELEASE\\pathmodelr_0.12.tar.gz", repos= NULL, type="source")
First, install the package using devtools: (install devtools first if you haven't already)
install.packages("devtools")
library(devtools)
Then the latest pseudo-stable version of the package can be installed directly using the following command:
devtools::install_github("GeertPostma/pathmodelr", ref="0.12Release")
Alternatively, the ref can be changed to any Git branch to test a different release. v0.12Release was used for the foundational Process PLS paper.
To get a rough grasp of the possibilities and to check if the installation was succesful, it is advisable to re-run the Val de Loir analysis. The full script for this analysis is available at https://data.mendeley.com/datasets/9x9h7fr4kn