search

PalEON-Project / Composition_Climate

Shifts in forest tree genera since EuroAmerican settlement.
MIT License
0 stars 1 forks source link

a few thoughts #5

Closed paciorek closed 8 years ago

paciorek commented 9 years ago

Hi Simon, I just took a quick glance at Composition_Ecology.pdf - is that what we are supposed to look at that you mentioned in the phone call?

I'm having a bit of trouble following based on what is in there at the moment but happy to provide feedback as it develops. It seems that you are estimating the density of a given climate variable for the locations where a given taxon is present. And then somehow you are able to just look at land use or just at climate change and get density estimates that include only one of the two factors and then compare how different the densities (e.g. is the land-use-only density more similar to modern than the climate-change-only density) are based on Hellinger distance between the densities. Seems like a useful way to go based on my limited understanding at this point.

SimonGoring commented 9 years ago

Yep, you've got it. Obviously it can be clearer, but we're partway there :)

Part of the reason I want to establish thresholds for the posterior means is that ultimately I'm transforming grid cells here to either 0 or 1 to generate the density estimates, so all the posteriors (as it is now) would be coded as 1. Anyway, I'm going to put some work into this paper next week, and when I've got a clearer draft and some leading figures I'll get back to everyone. Thanks for having a look over it though.

If anything jumps out please make an issue and I'll deal with it.

On Fri, Jul 31, 2015 at 8:17 PM, Christopher Paciorek < notifications@github.com> wrote:

Hi Simon, I just took a quick glance at Composition_Ecology.pdf - is that what we are supposed to look at that you mentioned in the phone call?

I'm having a bit of trouble following based on what is in there at the moment but happy to provide feedback as it develops. It seems that you are estimating the density of a given climate variable for the locations where a given taxon is present. And then somehow you are able to just look at land use or just at climate change and get density estimates that include only one of the two factors and then compare how different the densities (e.g. is the land-use-only density more similar to modern than the climate-change-only density) are based on Hellinger distance between the densities. Seems like a useful way to go based on my limited understanding at this point.

— Reply to this email directly or view it on GitHub https://github.com/PalEON-Project/Composition_Climate/issues/5.

SimonGoring commented 8 years ago

Paper is submitted & accepted with revisions. Closing this issue.