main driver 2 - spatial occurrence

tmesaglio commented 2 years ago

@wcornwell @eflowerproject I still need to do a bit of minor cleaning of points randomly in the middle of the ocean, but here's the heat map (with thanks to Sophie for her awesome code).

Looks pretty good!! Three hotspots = the Top End from Broome to Arnhem Land, SW Western Australia, and the Daintree

This is at a resolution of 1 degree x 1 degree cells

wcornwell commented 2 years ago

maybe try this: https://www.rdocumentation.org/packages/CoordinateCleaner/versions/2.0-20/topics/cc_sea ?

tmesaglio commented 2 years ago

maybe try this: https://www.rdocumentation.org/packages/CoordinateCleaner/versions/2.0-20/topics/cc_sea ?

yeah I tried that (I used it to clean a few records that had coordinates in herbaria), but unfortunately removing sea records also removes all of the island records; essentially it only treats continential Australia + Tasmania as 'land', and everything else like Xmas, Norfolk, Lord Howe etc get wiped out too.

Should be pretty quick to fix manually

wcornwell commented 2 years ago

also, i'm usually not in favor of this, but this might be a good situation to split the continuous variable into bins e.g.

0 unphotographed species known
1 unphotographed species known
2-5
5-10
10-100
>100

might make for easier interpretation by readers compared to the continuous scale--for example it's good for people to be able to see the difference between 0 and 1

tmesaglio commented 2 years ago

Yep no worries, can do. With those exact bins?

wcornwell commented 2 years ago

idk, might have to fiddle with the bins, but i think the difference between 0 and 1 and 1 and 10 are getting lost currently.

tmesaglio commented 2 years ago

@wcornwell new version with binned data (ignore the legend formatting, I have to fix that; the colours are correctly assigned to each bin, it's just the order that's been swapped around)