Intro to ML using R

[fpsom]

A first draft of the tutorial for the intro to ML using R. Content is currently the same as the course delivered at SIB (https://github.com/fpsom/2020-07-machine-learning-sib).

It will be adapted and trimmed down to fit 2-3 hours (as opposed to the two full days it's currently designed for).

[hexylena]

Fantastic! Is this all of the episodes or just a subset?

[fpsom]

Currently this contains all content from all episodes, merged into a single tutorial. Given the size, I'd be cutting stuff down as much as I can, to better fit the expected duration.

[hexylena]

Sounds good! We could also split them up into multiple tutorials? But I definitely understand if you just want to record something shorter for GCC :)

[fpsom]

Hm. Given that we do have the material, it may be possible to split this into multiple ones. However, as a first iteration, I'd argue to have a shortened (probably) version, in order to really check the time necessary to deliver.

After GCC (i.e. a first delivery of the content in this form), we should definitely revisit this. :)

[hexylena]

Works for me, let's do that :)

[hexylena]

We should import the images at some point, but otherwise looks ok to start

[fpsom]

I've made another pass at the material, and added also the images (hopefully in the correct path).

I think what is missing now is formatting, and another set of eyes to review content itself.

[hexylena]

@fpsom it seems you created this PR from the main branch, as a result I can't push back changes. I've got the formatting changes mostly done. Could you either 1) give me permissions on your fork to push commits, or 2) make a branch off of main with your changes, and then PR that branch?

[fpsom]

Thanks @hexylena ! I've just added you to the fork :)

[hexylena]

Ok, formatting looks better. There are some TODOs I've left in the tutorial for whenever anyone has time

[fpsom]

Thank you @anuprulez for the very positive and constructive feedback. I think I've addressed all your points in this new commit. :)

[anuprulez]

I guess some more work is remaining for the tutorial as the title suggests? Or WIP can be removed?

[fpsom]

Thanks @anuprulez ! I've just removed the WIP. I think the material is in a good enough state as they are. I do plan to make some updates (mostly in the solutions of the exercises), but I'm not sure that it'll be in the next few weeks.

[hexylena]

Yeah, there are a couple TODOs, but they can wait. The only one that should be fixed is probably:

TODO: (There is a lot of information to unpack here, so we will discuss during the course what these 5 lines of code actually do)

and those 5 lines aren't discussed afaict. (Is it even necessary?)

[fpsom]

Hm... I actually missed this ToDo...

Well, I don't really think that it would make sense to explain these here. If anything, a link to the Visualization in R tutorial could be included. Otherwise, I feel that this will boil down to a ggplot discussion.

But I'm open to suggestions - I can add a few sentences there, but (as I said), they might feel rather disjoined.

[hexylena]

Sounds good to me, linking is a good option

[fpsom]

Perfect - done! :)

[fpsom]

Hm... Did I do something wrong now by pushing another commit? Referring to the "dismissed anuprulez’s stale review" notice above..

[hexylena]

No, that's normal. Every commit you push after someone has reviewed clears old reviews to ensure that no unreviewed content gets merged. Everything must have a review as the last step

[fpsom]

Ah, clear now. Thanks @hexylena ! :)