In couthcommander/rcmoonpie: Redcap Data Dictionary

knitr::opts_chunk$set(echo = TRUE)
knitr::opts_chunk$set(R.options = list(width = 100))

Motivation

REDCap is a data collection tool. Data can be exported and read into R as a CSV file. Some basic manipulation (like setting variable labels) can be performed by downloading an R script also provided through the REDCap system. Parsing this R script can provide a data dictionary, which the package will provide in additional formats (YAML, CSV). This package will allow a user to modify the data dictionary which will update the R script used for importing the REDCap data.

Functionality

Use script2info to build a data dictionary (list) from the REDCap provided R script. While you can examine the list created (for example with str), you will likely want to use the dd2df function to convert the list to a data.frame.

library(rcmoonpie)
dd <- script2info(system.file("examples", "ex_script.R", package = "rcmoonpie"))
dd2df(dd)

You can export your data dictionary to YAML or CSV with the dd2yaml and dd2csv functions. Set the "file" argument to a file location, or alternatively allow it to print to standard output.

YAML and CSV can be imported back into a data dictionary with the yaml2info and csv2info functions.

dd2yaml(dd)
td <- tempdir()
dd2csv(dd, file.path(td, 'test.csv'))
dd_alt <- csv2info(file.path(td, 'test.csv'))
# exporting to CSV and back will have equivalent objects
all.equal(dd, dd_alt)

Variables and factor levels can be removed with excludeVar and excludeLevel. Excluding can be undone with unexcludeVar and unexcludeLevel. Note that you can also accomplish exclusion by modifying YAML or CSV output.

dd <- excludeVar(dd, 'data', 'redcap_event_name')
dd <- excludeLevel(dd, 'data', 'sex', 666)
dd2df(dd)

Rather than passing excludeVar a variable name, a regular expression pattern can be used to exclude all variables matching the given pattern. You can check which columns a pattern includes with findVariableByPattern.

# which variables include an underscore character
findVariableByPattern(dd, '_')
# exclude all variables that have an "_" in the name
dd2df(excludeVar(dd, 'data', '_'))

The final output will be a new R script to replace the original downloaded from REDCap. It will reflect any changes to the data dictionary. Use the dd2script function to create the R script. Like dd2yaml it has a "file" argument that can be set to a location or left blank. It also has an argument "factorHandle" that can be used to change factor variable behavior. This can be set to one of three values:

• duplicate - The original character string variable will be copied to a new factor variable that includes ".factor" at the end of the variable name. This is the default.
• unchanged - The original character string variable will not be turned into a factor variable.
• changed - The original character string variable will be turned into a factor variable.

dd2script(dd)
dd2script(dd, factorHandle = 'unchanged')
dd2script(dd, factorHandle = 'changed')

couthcommander/rcmoonpie documentation built on May 17, 2022, 12:35 a.m.