first use of function with side effects needs explaining?

[steltenpower]

On
levels(respondent_floor_type)[2] <- "brick"
a student said:"if you're not putting the result of the levels function into a variable, what's going on?"

[atheobold]

@steltenpower -- I would agree that this is a bit hard to understand for beginners. Personally, I would like to see this changed to some of the functions from the forcats package, such as fct_relevel(). Would you be interested in putting in a pull request improving this section of the workshop content?

[cgross95]

If there is still any interest in making these changes throughout the episode, I would be happy to contribute a pull request!

If so, the same section uses, e.g., in line 395
respondent_floor_type <- factor(respondent_floor_type, levels = c("earth", "cement"))
to reorder the levels. If the forcats functions are preferable, would it be worth it to change this to
respondent_floor_type <- fct_relevel(respondent_floor_type, c("earth", "cement"))?

Since these commands use pretty much the same syntax, I would also understand if this change unnecessarily introduces the fct_relevel function and it would be better to leave it out.

[juanfung]

@cgross95 that would be great! Since we have not introduced forcats yet, I would prefer the former but if you are feeling ambitious perhaps we could have a callout with a quick intro to forcats illustrating the alternative, and maybe some other uses. Either way, I'm happy with the explicit use of factor as in line 395 for consistency.