Dataset candidate: world innovation data

[mtennekes]

I came across another potentially interesting dataset, the world innovation index. It is an index based 7 pillars, from which the first 5 are related to input (what does the government in order to increase innovation), and the latter 2 to output (what innovation does the society create).

There is a yearly time series from 2013 to 2019:

Also interesting is the difference between 2019 and 2013:

As you can see, there are quite some countries without data. Let me know if this could be useful for the book. If so, I will clean up the script and make it reproducible.

[Nowosad]

I like this dataset. I agree that having this number of missing values could be seen as an issue, but, on the other hand, it is often the case when working with real data...

[mtennekes]

Done! See https://github.com/r-tmap/tmap-data/blob/master/R/02-prepare-world-innovation.R

This data is a 3d array: country iso3 code x pillar x year. Which can be cast into a vector data stars object.

I've used wm (used to create world_all) as shape to join the data with, with a better coverage than tmap's World; still a few missing values, but less. Just discovered the countrycode package. Highly recommended!

Currently, tmap only supports visualization of a slice of vector data cubes (see https://github.com/mtennekes/tmap/issues/458).
Ideally, it should be possible to do tm_facets(by = c("year', "pillar")).

Where shall we store these datasets? In the book repo, via a data-package, or ...?

[Nowosad]

Awesome @mtennekes !
Regarding the question... I have some ideas - we can discuss them during our next meeting.

[Nowosad]

We can decide it later.
For now - we should keep the data in the book repo. In the future, the data can either go to the tmap repo or some new one (e.g., tmapdata.