Readme for Dropout Meta-analysis

Introduction

This readme describes the files and data for the dropout meta-analysis.

To reproduce the results:

  • Make sure all files listed below are in the same directory (HTML files can be omitted)
  • Open dropout.Rmd
  • Knit the file

Note that re-kniting the files may potentially overwrite the output files in the folder.

Note that the knitting needs to occur from dropout.Rmd all at once. Knitting chunks in dropout.Rmd that call on child .Rmd files such as substances.Rmd fails when run interactively.

Files

The following table describes all of the files in the directory. The output in the *.html files includes information about the system that was used for for the latest run of the dropout.Rmd file.

File Description Output
categorical.Rmd Knitr child file.
Is called from dropout.Rmd.
Evaluates study-level categorical variables.		 Returns the knited analyses of study-level categorical variables, including plots and diagnostics as specified in dropout.Rmd.
Depends on the values in the parent environment created by dropout.Rmd.
continuous.Rmd Knitr child file.
Is called from dropout.Rmd.
Evaluates study-level continuous variables.		 Returns the knited analyses of study-level continuous variables, including plots and diagnostics as specified in dropout.Rmd.
Depends on the values in the parent environment created by dropout.Rmd.
Data Dictionary.xlsx Excel file that includes the variable names, their descriptions, and their possible values, in human-readable format. NA.
datadictionary.R R code that facilitates translating the variable names in the Meta Dataset file into human readable names embedded into the output generated rom the *.Rmd files. NA. When called, it creates a series of matrices that are used by the other *.Rmd files.
diagnoses.Rmd Knitr child file.
Is called from dropout.Rmd.
Evaluates the study-arm-level comorbid diagnoses variable.		 Returns the knited analyses of the study-arm-level comorbid diagnoses variable, including plots and diagnostics as specified in dropout.Rmd.
Depends on the values in the parent environment created by dropout.Rmd.
dropout long.html Output from dropout.Rmd using the long=TRUE option. This includes more of the code, diagnostics, profile plots, and other inputs and outputs. NA. It is output. Inclusion of plots is dependent on setting plots=TRUE.
dropout short.html Output from dropout.Rmd using the long=FALSE option. This includes less of the code and diagnostics, excludes profile plots, and limits the printing of some inputs and outputs to make it more easy to evaluate NA. It is output. Inclusion of plots is dependent on setting plots=TRUE.
dropout.Rmd Parent knitr file.
Includes the parent rmarkdown and loops for calling the child knitr files.		 Outputs the synthesis of everything from dropout.Rmd and the child knitr files.
Output saves as dropout.html, and is manually renamed to one of the dropout long.html or dropout short.html filenames.1
functions.R All of the custom functions, convenience functions, and function wrappers used in the *.Rmd files. Depends on the functions called.
Meta Dataset.xlsx The input dataset for analysis.
Includes the data dictionary.		 NA: data set.
readme.html The output from readme.Rmd. NA: it is output.
readme.Rmd The file that constructs the readme information. readme.html
substances.Rmd Knitr child file.
Is called from dropout.Rmd.
Evaluates study-arm-level substance use and use pattern variables.		 Returns the knited analyses of study-arm-level substance use and use pattern variables, including plots and diagnostics as specified in dropout.Rmd.
Depends on the values in the parent environment created by dropout.Rmd.

  1. Programmatically kniting dropout.Rmd to rename the html output ran into errors with duplicate chunk names from the knitting of child files. The code therefore outputs by default to dropout.html based on the .Rmd filename, and the dropout.html file is manually renamed.↩