I'll be gone next week, so let me give you the third weekly assignment now. Look closely at the structure of https://github.com/wibeasley/RAnalysisSkeleton and try to understand at least 80% of it.

• Try to identify themes like relative/working directory.

• In your head, describe the pros & cons of saving data as csv vs rda files. Identify when I favor one over the other. Make some guesses why.

• For now, don't worry about any of the files in the Report1 directory. That's for the next week's assignment. This week is mostly about reading, cleaning, & persisting data.

• Fork this repository, pick a different dataset, and do something that mimics what I did with the cars dataset. Download a 'BreastfeedingSleepFake.csv' CSV in the DeSheaToothakerIntroStats repository. You should have received an invitation to the repository a few minutes ago.

  This is currently a private repository that contains code for an upcoming textbook. You may have met Larry Toothaker in the department. He's semi-retired now, but taught me and @bard1536. And Lise DeShea was sitting in front of you at SCUG two weeks ago.

Some of this & next week's assignment is covered in the May 2014 SCUG presentation.

Read about knitr, and try to understand the Report1.Rmd file. The only real difference is that most of the analysis code is split into a separate R file (ie, Report1.R). The Rmd hooks into the R file with the line:

 #This allows knitr to call chunks tagged in the underlying *.R file.
read_chunk("./Analyses/Report1/Report1.R")

It will be helpful to focus on the chunks (and how they're linked across the R and Rmd files). Pay attention to lines in the R file that start with ## @knitr......

The fourth assignment continues with RAnalysisSkeleton, but focuses more on analysis and reporting. I'll see how the previous two assignments go before I get more specific here. At minimum, we'll cover the Report1.R file.

We might jump into knitr and the Rmd file too. If so, an important piece to understand in the Rmd file is the line:

read_chunk("./Analyses/Report1/Report1.R") #This allows knitr to call chunks tagged in the underlying *.R file.

@bard1536, will he be doing anything with knitr anytime soon in your plans?

Issue list:

• I don't actually have any issues, I just wish I was more mysterious.
• Also, I need this here for the assignment.

In the spirit of #1...
For the second assignment

• Install the REDCapR package on your local machine. See the 'README' file for directions.
• Create a new R file in the playgrounds directory. Have it use at least one REDCapR function. Files in this directory aren't officially incorporated into the package. They're mostly for our developers to try out some things before integrating new code into the package.
• Commit the new file to your 'fork', and then submit a 'pull request' to me.
• If you see any mistakes or misspellings, don't be afraid to create a quick pull request that fixes it. Sometimes it's as easy as doing it in the GitHub browser. In my mind, even little small corrections in the documentation are helpful (for example, see my two tiny ones in this package that I use all the time), because it reduces confusion for future users who may have gotten derailed. If they're not what the main developer wants, it's easy to avoid pulling it.

Most of the things in this assignment and others will have lots of examples online. If you find you're stuck on any one thing for more than ~10 minutes, email or call me. One of @bard1536 's motto goes something like, "its ok if you're bruised, but call before you get bloody."