OZ FETP: Introduction to data management

[zkamvar]

https://github.com/ArminderD/Data_management

This exercise was developed for the Australian Field Epidemiology Training Program, The Masters of Philosophy (Applied Epidemiology) offered through the Australian National University. This exercise provides an introduction to the RStudio environment including basic commands and the generation of basic graphs. The exercise also serves as an introduction to data management in epidemiology introducing the concepts of data checking, data cleaning, and manipulation of variables.

[zkamvar]

Preview document: https://htmlpreview.github.io/?https://github.com/ArminderD/Data_management/blob/master/Data_managment.html

Because there are questions and potential solutions, this will be classified as a practical instead of a Case Study.

Some issues to address, though some of these need to be cleared with the authors of the material beforehand.

• Remove attach(). There is never a good reason to use this function.
• Convert image tables to text
• Find out why there is an extra column "gram" in the data set
• Standardize code formatting
• Standardize markdown formatting (e.g. Graphics is a level 4 header, but should be something like a level 2)
• Standardize language used (e.g. there are places which mentions a graphics menu, but this doesn't exist in RStudio)
• Standardize data subsetting (e.g. sometimes the user is instructed to use x[x$a > 1 & x$b < 40] and sometimes they are instructed to use subset())
• Change ifelse cascade to dplyr::case_when()
• Update factor manipulation (they don't need to be converted to numbers first)
• Update dates section to have users not convert dates to character
• Consider using tidyverse functions in place of merge and reshape (maybe?)