Book Outline

[christophsax]

@jlivsey also move the discussion here.

I put the book outline in a separate file, and added some stuff to it: https://github.com/christophsax/x13book/blob/master/proposal/outline.md

It may also make sense to focus on the outline first. If that is clear, we the other questions are basically answered.

Some ideas:

• I think we could build the book around a cheat sheet with references to the chapters. Probably would make sense to have an idea of the cheat sheet at an early stage.

• As discussed before, each section discusses a specific practical problem. We could also add a 'Case Study' box with a more challenging question to each section. Ramadan was the only thing that came to my mind, but I am sure will will find other great examples. I remember one of the Eurostat people talking about a Tunisian time series where they had strong tourism driven Easter effects along with Ramadan. I think it would just make it much more interesting to see these cases instead of AirPassengers all along.

• So we could have one (or a few) interesting standard series (e.g. US Unemployment) which we use throughout the main text, and use the exotic series in the case study boxes only.

[jlivsey]

I agree that finalizing the outline is a good strategy. To your points:

• I love the idea of having a cheat sheet. Building the book around it might be a bit much. I would like to think of the cheat sheet as being built along with the text.

• Again, love the idea of a 'Case Study' go accompany each section. I have many interesting retail series we could include here. One example is the apple cider retail sales has an interesting outlier structure and seasonal pattern.

[christophsax]

Building the book around it might be a bit much. I would like to think of the cheat sheet as being built along with the text.

Yes, let's do that 😄

So just add a note to outline.md, whenever you have a good idea for a case study.

What do you think of the structure in outline.md? What do we miss?

[jlivsey]

I like outline.md so far. I had the idea to link each data problem with a spec. This would improve functionality of the text. Then a user could ask:

'how do I add outliers?' and be directed to the outlier data problem chapter or 'how do I use the outliner spec?' and be directed to the outlier data problem.

I took a first crack at incorporating this simply into outline.md.

[christophsax]

Not too sure if I like it. I would keep the structure as flat or linear as possible. 10 to 15 chapters, divided in 3 to 5 structuring parts makes a good reading, I generally don't like hierarchical over-structuring with subsections and sub-subsections.

Surely, holiday effects and trading days are technically part of the regARIMA model. But it seems almost irrelevant to me from a user perspective, who has either a trading day or a holiday problem, and does not care too much about the spec involved in solving it.

Following the spec structure also does not make too much sense logically. Outlier have their own spec, because it is needed for detection, but final outlier adjustment is done in the regARIMA as well, so why having them in a separate chapter? Solving a specific problem may involve working with several specs.

"How do I use the outlier spec?" is also not really the question we want to answer. A spec is just a tool to solve a problem. 'how do I add outliers?' is better, but the questions start with: Why are outliers are problem? Why should I add them? How do I add them? And answering these questions probably involves multiple specs, e.g., some diagnostics.

I would quickly cover regARIMA in the beginning as part of a general overview on the working of X-13, pretty much dissolved from concrete data problems.

Bottom line: I don't think that we should use the X-13 specs as the structuring element of the text. Trading days, Holidays, and Outliers each deserve its own chapter. Structuring by the specific problem is also what I liked so much in the ONS handbook.

[jlivsey]

This works for me too. Honestly, it was just an idea but certainly makes the book less flat (more subsections). I'm completely fine with that perspective.

Let's roll outline.md back to the previous commit and I'll take another stab at it from this more linear/flat perspective. I think it will change my comments...

Good conversation!

[christophsax]

I got a seasonal adjustment question and thought we could start collect topic here:

https://github.com/christophsax/x13book/blob/master/topics/existence-of-seasonality/existence-of-seasonality.md

The specific question was whether or not to adjust some series. Our book should give some good answers to that.

So I just wrote down what came to my mind. No intention to have the specific example in the book, but it may help us to better see the practitioner's problem.

The specific case might be somewhat interesting, as the QS Tests and the M7 Tests point to a different answer, I guess.

But more interesting is to have a blueprint for other topics. I like 'spin' R files that render automatically into a easily readable .md Document on GitHub.

[christophsax]

Wanted to open a fresh issue, but somehow parts of GitHub do not work as they should right now.

[christophsax]

Link above now here:

https://github.com/christophsax/seasonaldaily/blob/main/topics/existence-of-seasonality/existence-of-seasonality.md

[christophsax]

Closing the outline discussion, we are good!