Subsetting Notes (and one note on summarization)

[avahoffman]

• - I wonder if talking about special characters in names in this lecture is too soon? Maybe only address it if it comes up to be "just in time" - https://jhudatascience.org/intro_to_r/modules/Subsetting_Data_in_R/Subsetting_Data_in_R.html#25
• - can we make this more concise? - https://jhudatascience.org/intro_to_r/modules/Subsetting_Data_in_R/Subsetting_Data_in_R.html#28 - maybe just the sentence "Replace any curly quotes ‘, “ you paste with ',"" .. could also move this to later if we discuss special characters later.
• - Why is the first column showing up all in green?
  this kinda threw me off

[carriewright11]

• slide 36 has the header for the next section on it (subsetting columns) https://jhudatascience.org/intro_to_r/modules/Subsetting_Data_in_R/Subsetting_Data_in_R.html#36

[carriewright11]

• could simplify some of the slides maybe, although repeating is also useful
• maybe move the illustration slide to the end about filter before lab 2
• could probably remove this slide: Subset columns of a data frame: dplyr where it shows select and pull together

[avahoffman]

I still think relocate and order might be pretty specific use case. Perhaps it could be a single slide ("Check out these functions if you want to rearrange columns"). In general, select can reorder things so maybe just encourage them to do that?

[carriewright11]

I actually kinda prefer teaching only relocate and not everything() I think it is easier - what do you think about that @avahoffman ? (it is quite hard/annoying to move a column to the middle without relocate, which I find I do quite often if I make a new column based on a column in the middle).

Happy to move order to extra slides

[carriewright11]

(noting that we use relocate in the lab)

[carriewright11]

• need a comma between Chambers AL in the last question of the lab

[carriewright11]

I noticed that the summarization slides use tb data with columns that have numerical names and require backticks around them - maybe we should use clean_names on those or use a different dataset if we decide to drop that discussion?