datascience-book / textbook Goto Github PK

Should graphs all be in ggplot or is it ok to use base graphics? I realized I was doing everything in base as that's what I usually use during the modeling process, but figured I should ask

In the load and I/O section, we should cover the basics of read_csv and read_excel. Then cover read_dta from haven to help econ/soc sci students transition. Should we also cover JSON and XML as those are more common when working with APIs or save that for later?

In the data manipulation section, I need help thinking through the string functions. Do we want to teach both base R (gsub, grep, grepl, regexpr, etc) and stringr (str_replace, str_count, etc)? Or maybe lean more on stringr and have a table showing the similarities? What do you think?

Hey guys, what would you think about adding a section at the end of each chapter called "Best Practices?" It could cover a more practical aspect related to the chapter's material, for instance, in ch. 3 I added a section about comments. It may also be better to put it all together in a separate chapter. Thoughts?

What do we want to do with hypothesis testing? Assume knowledge? Refresher? In-depth treatment?

How much of math and stat should we assume that our audience understands? E.g., should I assume they know correlation, variance/standard deviation, and hypothesis testing?