input formats

[mpadge]

input formats

[mdsumner]

Re other formats, geosphere has two modes p1[i] to p2[i] (paired) or sequential along p1. Can you sensibly get all pairwise distances in one function that way, include segment = T/F, x-only for sequential?

geosphere doesn't do all pairwise anywhere (like spDist), does it?

Definitely avoid any classes, matrix long-lat is fine.

[mpadge]

And I think we can exclude sf objects because:

1. If people are using those, then they've likely loaded sf along with all it's internal distance calculation abilities; and
2. It would just open up a minefield of decisions regarding between which objects distances ought to be calculated

[mpadge]

closing now and deferring to #4 for discussion of geosphere-type implementations (and yeah nah, geosphere only does sequential, absolutely no pairwise).