Reviving the R syllabus - 2026

[colinleach]

A lot of content was written back in summer 2023, but most of it was never merged.

The documents now need updating in two main ways:

• Incorporate learning from other tracks, especially the recent Julia syllabus.
• Add references to Tidyverse functionality. This collection of packages was added to the test runner in late 2023, after work on the R syllabus was suspended.

See also issue #412, which is mainly about adding hints.

A lot of draft content is on my repo.
We'll need to move this to PRs as it gets updated.

Concepts and exercises already merged, but needing updates:

• Basics
  • Lasagna
• Booleans
  • Annalyn's Infiltration
• Conditionals
  • Cars Assemble
• Vectors
  • Elyse's Enchantments

Concepts and exercises with unmerged drafts:

• Vector Filtering
  • Elyse's Analytic Enchantments
• Vector Functions Vector Functions concept #430
  • Bird Watcher Bird Watcher concept exercise #453
• Strings Strings concept #458
  • Log Levels Log Levels concept exercise #459
• Switch Switch concept #433
  • Blackjack Blackjack concept exercise #461
• Loops Loops concept #440
  • Mixed Juices Mixed Juices concept exercise #441
• Set Operations
  • Ozan's Playlist Ozan's Playlist concept exercise #435
• Nothingness
  • Name Badges Name Badges concept exercise #449
• Randomness
  • Captain's Log
• Regular Expressions
  • Regular Chatbot
• Errors Errors Concept #451
  • Factory Sensors
• lapply and sapply
  • ??
• names Attribute Names attribute #436
  • named-resistor-color Add concept exercise named-resistor-color #446
• Lists (Lists concept #437)
  • Need For Speed Need for Speed concept exercise #448
• Matrices and Arrays
  • ??

[depial]

Incorporate learning from other tracks, especially the recent Julia syllabus.

My romp through the type system of Julia has made me quite a bit more interested in this aspect of R which is why I've floated the idea of adding more detailed material on vectors. I'm not sure if I've miss interpreted things, but it's very interesting because it seems like vectors are the supertype of most types in R. For example, with logical:

length(FALSE)            # has length
# => 1

FALSE[1]                  # indexable
# => FALSE

for (v in FALSE) print(v) # iterable
# => [1] FALSE

typeof(c(FALSE)) == typeof(FALSE)
# => TRUE

c(FALSE) == FALSE
# => TRUE

is.vector(FALSE)
# => TRUE

So FALSE seems to be a single element vector (a.k.a "scalar") under the hood and the other types integer, double and character behave the same way. Unsurprisingly, NULL functions as the null vector. More surprisingly, more complicated objects also return TRUE with is.vector(), like list, matrix, dataframe etc.

Let me know if I'm on the wrong track and have misunderstood something, but if you want to include some information about this, I think simply putting passing mentions in the docs might be good enough that it would slowly help build an intuition for the type system (I'd guess an overt discussion is probably to much information for beginners). Of course this aspect could just be ignored (as I think it was in most of the learning materials I used), but I think it would be nice to at least allude to it in a non-intrusive way.

[colinleach]

I think we're talking classes (in an OO sense) rather than types internally, but that doesn't detract from the "everything is based on vectors" idea. Now we just need to find ways of gradually hammering it home in the syllabus.

[depial]

I think we're talking classes (in an OO sense) rather than types internally

Oh my... After so long of trying to censor the words "class" and "object" from my vocabulary, it looks like I'll have to reintegrate them at some point 😆

[colinleach]

Worse than that. We both need to try and understand the difference between S3, S4 and R6 classes. Fortunately the OO details are mostly hidden from end users, though very important for people writing libraries.

So: R is OO internally, imperative in (traditional) common use, and increasingly functional in modern use. What the term "multi-paradigm" is invented for!

[colinleach]

Is it time to start merging some Concept updates? I've been through the above list as far as Vector Functions to add a bunch of reference links, plus some other updates. The next stage is to add some more comments about vectors to the early concepts. I just did that for Basics, which may be ready to PR.

All the about.md files have changed, but also some introduction.md and links.json. It might be simplest to overwrite everything from stuff in my repo, and let Git sort out what changed.

[colinleach]

I just updated Booleans, so I think that is now ready for a critical eye to look at it.

[depial]

Sounds good! I'll be busy for the next couple of hours, but I'll try to take a look when I'm free later.

[depial]

Is it time to start merging some Concept updates?

Looks like I'll be able to properly review PRs here now, so you can open at will!

[depial]

Is this some weirdness going on in Conditionals? (I haven't completed either exercise as of yet)

[colinleach]

We need to remove Cars, Assemble. I should never have created it, but I was (even more) ignorant in those days.

There will be another thing to sort out with Vector Filtering. I drafted both elyses-analytic-enchantments and bird-watcher, but fortunately neither was merged. We'll need to choose one and forget the other.

[depial]

I assume by remove Cars, Assemble you mean remove Vehicle Purchase? I'll open a PR to deprecate it.

[colinleach]

Fortunately, you can read my meaning better than I can write it!

[depial]

Well, that cleaned right up!

[colinleach]

Continued in #492.