Graduated colors showing "out of range" incorrectly? Large polygon dataset

518
6
01-23-2022 06:48 AM
Labels (1)
CraigJariz
New Contributor II

Hi Community

I have a polygon dataset with ~610,000 square cells that I'm trying to apply a graduated color symbology to. When I do so, a large portion of the dataset shows as "out of range" even when I believe I have adjusted the intervals appropriately. See attached image: in this example I've selected one of the "out of range" cells that has a value of ~766, when my symbology goes from 730-795 as shown.

I hunted a bit for this but could not easily find someone with similar issues, so here I am. Is this too many cells for this symbology type (I can't imagine that's the case, but it never hurts to ask). Am I doing something else wrong? I've fiddled with this for an hour and am absolutely stumped.

Your thoughts appreciated

 

Weird_Symbology.PNG

Tags (2)
0 Kudos
6 Replies
DanPatterson
MVP Esteemed Contributor

change it from "manual interval" to another option to see what happens.

I suspect that you have too many polygons, it appears that the one you are clicked on has an area of 100 which suggests it looks like a raster 10x10 unit cell size. and there would be way too many polygons to symbolize for your area


... sort of retired...
0 Kudos
CraigJariz
New Contributor II

Oh, it has the same issue for other interval options too; for example if I do equal interval or defined interval it doesn't end up picking up portions of the range by default, so I've been using manual interval to make sure that it does.

If it's really that there's a hard cap on the number of polygons the program can read in when setting up "graduated colors" symbology, that seems like a weird limitation. It can display all of them with "single symbol" or "unclassed gradient" and load the whole data table just fine, so why would it be limited for other types of symbology?

0 Kudos
curtvprice
MVP Esteemed Contributor

Sure you don't want to create a raster version of your square cells? It would display dramatically faster and you would not have this problem.

DanPatterson
MVP Esteemed Contributor

I agree with Curtis, that raster would be the way to go, or you could try a Dissolve, the reduce the number of polygons since it would merge adjacent "cells" which share the same value. and hence, reduce the number of polygons


... sort of retired...
CraigJariz
New Contributor II

Thanks all

I've looked at a raster for this but Pro struggles with Polygon to Raster with this dataset, I tried it once and it took 5 hours and this is a fairly nice PC.

I've got to take about 3000 similar polygon datasets (for different simulations) and merge them into a timelapse, so that would be a hurdle even if I did 5-6 at a time in parallel. I was hoping to just figure out a standard symbology that I could copy from polygon dataset to polygon dataset.

I'll keep fiddling.

0 Kudos
curtvprice
MVP Esteemed Contributor
I can't imagine why this would take five hours. What parameters are you using? Nevertheless if you rasterize on sequence number you only have to do it once and then you have a raster you can display with any joined value (no need to copy the dataset 300 times).

0 Kudos