Missing hints for concept exercises.

[depial]

It appears the hints.md files for the concept exercises are just blank stubs. This includes the four current exercises:

• annalyns-infiltration
• cars-assemble
• elyses-enchantments
• lasagna

[colinleach]

Yes, priorities have changed since these were drafted. Hints are now strongly recommended for concept exercises.

I think there are more changes needed, now that Tidyverse packages are available in the test runner:

• Update the introduction.md if necessary (though the concept about.md will need bigger changes).
• Check if the exemplar is still good, or should be using newer functions.

[colinleach]

Starting at the top, with lasagna:

• Would you like to write some hints?
• I made some small changes to the concept introduction.md here. I think it would be worth copying that to the exercise, in the same PR.

My guess is that these hints will look quite a lot like the ones you wrote for Julia.

[depial]

I'll take a look at making these changes and open a PR.

My guess is that these hints will look quite a lot like the ones you wrote for Julia.

I'd also imagine they won't differ by too much since this is still largely the most basic functionality, but I'll try to keep my eyes open for idiomatic differences too.

[colinleach]

For booleans and annalyns-infiltration, the Introduction is extremely short. I think it would be best if I expand it to resemble the Julia version.

Update: the new version of the Introduction is here if you want to copy it to the exercise.

[colinleach]

For conditionals and cars-assemble, I did a significant rewrite on the docs. The new introduction is here.

I'm not satisfied: either that intro contains too much about ifelse() and if_else(), or we need an exercise that gives students chance to use one or both of the short forms. Do you see an easy way to add a relevant task to cars-assembe?

[depial]

I feel like one of the most important sides of ifelse() is its use with list vectors, so I'm not sure it needs to be introduced here. I can't think of specific example, but I know its vectorization has tripped me up in the past.

The dplyr version seems to have even stricter usage (in terms of missing values and/or type stability).

I would possibly wait to introduce them until vectorization could be talked about.

[colinleach]

Suggestion: remove ifelse from conditions/introduction.md and the exercise, leave it in the about.md as a forward reference.

The dplyr version seems to have even stricter usage

You'll see that a lot. A philosophy of Tidyverse is that the base code doesn't complain enough, so bugs slip through. Next level: tibbles replace dataframes under the motto "do less, complain more". It's all part of taking a scripting language hacked together by grad students (mostly in social science departments), and professionalizing it with modern software engineering practices. An ongoing project...

In terms of a 'vectorization' concept, that's pretty much what sapply-lapply is about. We could expand that.

[depial]

Suggestion: remove ifelse from conditions/introduction.md and the exercise, leave it in the about.md as a forward reference.

This seems like a good compromise. In the forward reference, I would probably make sure to mention that its a vectorized version of if-else, even it that could sound like gibberish to some students.

You'll see that a lot.

Good to know! I've learned a bit about the tidyverse, but I'll admit that I avoided using R in any instance I could have used Pandas, simply because I felt more proficient in Python. Oddly enough, that probably makes my (limited) knowledge in R be quite Base-R biased from doing a bunch of coding exercises :D