Use #1: Targets Pipeline

The full use of the demcrit package relies on patient data that cannot be shared publicly for privacy reasons. However, the package can be executed on the data owner’s computer (Laboratory of Neurosychology at the First Faculty of Medicine, Charles University, and the General University Hospital in Prague).

Set-up

Before starting, ensure that you have a local installation of the demcrit R package. If not, see the package overview for installation instructions, or visit the package website’s reference list for function documentation.

Load the package if you wish to access the documentation directly in RStudio:

library(demcrit)

Next, clone the demcrit repository to a local folder of your choice (here referred to as studyfolder), and install the necessary dependencies:

devtools::install_deps()

Ensure that all required data files are present in the studyfolder/raw-data directory.

Data Checking

The pipeline first runs the data_paths() function, which checks for missing files and naming inconsistencies:

help("data_paths")

Additional functions help prevent data-entry errors or structural inconsistencies before analysis:

help("check_compatibility") # Looks for incompatibilities between raw item data specific for this project and REDCap data from a larger database.
help("check_names") # Looks for inconsistencies in patients names in the raw item data compare to REDCap databes.
help("check_ranges") # Looks for impossible test scores.

Running the Pipeline

Finally, make sure that the apaquarto extension (version 4.4.1) is installed in the studyfolder/_manuscript directory.

You can then run the full pipeline (including manuscript writing) via:

#install.packages("targets")
#install.packages("tarchetypes")
targets::tar_make()

To visualize the workflow, use:

targets::tar_visnetwork() # full workflow
targets::tar_glimpse() # simplified workflow (targets only)