search

rociojoo / CranTaskView-Track

Cran task view project for tracking data
3 stars 0 forks source link

trajectories of what; geographic or cartesian #2

Closed edzer closed 4 years ago

edzer commented 4 years ago

You mention "organisms", "individuals", "biologging" and "animals", I would suggest to generalize this to "objects" except for cases where you have to get specific. The CTV is based on a manuscript in "Animal Ecology", but I don't see how (all the) packages focus on animals too, or can't be used for moving non-living things.

It would also good to point out whether (and which) packages assume 2-D Euclidean (Cartesian) coordinates, and which assume geographic (long/lat) coordinates. I see too many people who assume wrongly that the software does the right thing for their type of coordinates.

basille commented 4 years ago

Hey Edzer, regarding your second point above, isn’t it the role of each package, where relevant, to mention this explicitly? It would be a long and tedious task to review all packages with this particular point in mind — to me, that belongs to package documentation and/or warnings in R.

edzer commented 4 years ago

where relevant,

do you know of examples where it isn't?

In case you decide to not do this, it might be relevant to mention the risk, and that this is left to the users to verify.

basille commented 4 years ago

I did not think of packages for which it was not relevant, but I was thinking in terms of package development. Each package could certainly highlight it at the right time in the workflow (plus in the documentation).

I like the idea of a general warning in the CTV.

mdsumner commented 4 years ago

that this is left to the users to verify.

I agree with that at least, folks use Cartesian coordinates all the time in inappropriate ways. Is UTM ok? No. Is stereographic ok? No. When and where? Oh maybe that's ok ... It really matters.

I agree absolutely that it should be general, but beware of shooting from a basis that is extremely poorly held atm. "Spatial" is no help here.

rociojoo commented 4 years ago

I agree with the general warning.

rociojoo commented 4 years ago

The warning is in the text now (first paragraph): "Regarding (x,y), some packages may assume 2-D Euclidean (Cartesian) coordinates, and others may assume geographic (longitude/latitude) coordinates. We encourage the users to verify how coordinates are processed in the packages, as the consequences can be important in terms of spatial attributes (e.g. areas and angles)."

rociojoo commented 4 years ago

And the CTV is more general now. It refers to objects and individuals, and only when necessary, animals or humans.

rociojoo commented 4 years ago

And there is now an issue template to include new packages.

edzer commented 4 years ago

Instead of "areas and angles" I would mention "distance, speed and angles".

rociojoo commented 4 years ago

Done!

rociojoo commented 4 years ago

I'll close the issue if there is nothing more to add.