tucotuco
commented
8 years ago Nice to see something going into the idea phase so soon!
It might be too sophisticated and computationally out of range, but it might be interesting to visualize the whole continuum of presences to gaps (or have the option to turn this feature on) on a heat map, where the gaps are cold colors and the color in the gap is based on the distance to the nearest presence record within the distribution range.
On Wed, Mar 30, 2016 at 9:29 PM, Matthew J Collins <notifications@github.com
wrote:
Matt: When looking for gaps in data, is there a way to quickly reverse the density display on the portal map to have the darkest tiles be the ones with the least data?
Alex: The image kind of sucks.
Inverting dark/light it isn't the way to go... you just end up with an inverted heatmap which is still visually a density map. What you probably want to do, and I haven't figured out yet, is to invert the scaling function (i.e, go from logarithmic to exponential). It would probably also be worth playing with the color space to see what works best.
Also: this would be much more effective coupled with an expert range map or something . Obviously identifying the entire ocean as critically undersampled for pine tress is less than ideal.
Matt: Mmm. Yeah, I see what you're talking about. I like the idea of a mashup of range maps and occurrence data though. To the research ideas github project!
@tucotuco https://github.com/tucotuco - Watched your talk today, you mentioned this in passing.
— You are receiving this because you were mentioned. Reply to this email directly or view it on GitHub https://github.com/iDigBio/research-project-ideas/issues/6
mjcollin
Matt: When looking for gaps in data, is there a way to quickly reverse the density display on the portal map to have the darkest tiles be the ones with the least data?
Alex: The image kind of sucks.
Inverting dark/light it isn't the way to go... you just end up with an inverted heatmap which is still visually a density map. What you probably want to do, and I haven't figured out yet, is to invert the scaling function (i.e, go from logarithmic to exponential). It would probably also be worth playing with the color space to see what works best.
Also: this would be much more effective coupled with an expert range map or something . Obviously identifying the entire ocean as critically undersampled for pine tress is less than ideal.
Matt: Mmm. Yeah, I see what you're talking about. I like the idea of a mashup of range maps and occurrence data though. To the research ideas github project!
@tucotuco - Watched your talk today, you mentioned this in passing.