Add packaged DEM data

[fnattino]

(builds on top of #50, merge that first) Done

[fnattino]

GH actions are failing because the DEM data that I have added includes invalid pointers! It turns out SpatRaster objects cannot be easily serialized as rds/rda files:

"These classes hold a C++ pointer to the data and they cannot be directly saved to a ".Rds" file or
used in cluster computing. They cannot be recovered from a saved R session either. See wrap or
writeRaster to work around that limitation."
https://cran.r-project.org/web/packages/terra/terra.pdf

One can bypass this by using the dedicated saveRDS/readRDS functions. Can one make these work in the current packaged data setup? If not, we might as well store the DEM as a tiff file?

[cforgaci]

One can bypass this by using the dedicated saveRDS/readRDS functions. Can one make these work in the current packaged data setup? If not, we might as well store the DEM as a tiff file?

I did not know about this. If saveRDS/readRDS does not work in the current package structure, then the next option is to place a tiff into inst/extdata/. In this case, we need to indicate to the user that the dem needs to be loaded with system.file() separately as it is not made available automatically when the package is attached with library(CRiSp).

[fnattino]

Using terra::saveRDS works, but one would then "manually" call terra::readRDS to read the serialized object. A workaround is what I have currently implemented, which wraps the DEM into a PackedSpatRaster object that can be normally serialized/deserialized with the standard R package infrastructure. The drawback: one needs to call "terra::unwrap()" on the Bucharest DEM object before using it - but this is clearly stated in the representation of PackedSpatRaster objects:

> CRiSp::bucharest_dem
[1] "This is a PackedSpatRaster object. Use 'terra::unwrap()' to unpack it"

Is this an acceptable solution?

[cforgaci]

Is this an acceptable solution?

Yes, I think this is a good solution. Thanks!

[fnattino]

When updating the name of the dataset, I have noticed this function probably had a placeholder for the name of the dataset (st_write_list -> bucharest)? Also I made a suggestion to replace walk for walk2, which seems to simplify quite a bit the overall snippet?

[cforgaci]

yes, much better!

[cforgaci]

prima!

[cforgaci]

yes, much better!