Use simplified data set?

[kbroman]

When introducing R, we tend to get bogged down on dealing with missing values (NAs) and dealing with factors. And then dealing with both in read.csv(), via the arguments stringsAsFactors and na.strings.

For the "portal" data set, they both hit you hard with the sex column which has values "M", "F", or "" (blank).

Christina and I recently taught a one-day Data Carpentry workshop, and I taught dplyr and ggplot2 using the gapminder data. Those data were considerably easier because there are no missing values at all, and while there are factors, you don't need to do anything special with them.

We might consider using a reduced portal data set that has only the complete records, deferring discussion of NAs and factors to later or just skipping it entirely.

[byandell]

I like this idea of simplified portal data set to get quickly in to R. I also see value in the gapminder data, as there is an audience that largely uses categorical data. Not sure we are ready for two versions but I like what I see at https://github.com/kbroman/Workshop_DataCarpNSBE.

[sstevens2]

Notes from discussion:

• Clean up the data set so it doesn't have missing data
• Add info about what the dataset is and its history
• Put all the files into one zipped file for them to download
  • Including the file structure in this but have them explore this and add an exercise before where they put together a hypothetical project file

Solution: adding a lesson with this - could take about 30 min.