topic Re: Raster cell count changes in Spatial Data Science Questions

Raster cell count changes

Roosvan_der_Deijl — Fri, 22 Mar 2019 11:46:38 GMT

Hello there!

Disclaimer: I'm a GIS beginner, so this might be a stupid question, but I haven't found useful help yet on the internet.

Every time I do any calculation on my raster (raster calc, reclassify), the cell counts of the different values as well as the total cell count decrease by a factor (almost exactly) 4. The cell sizes and extent of the maps are the same.

This happens even when simply inputting the raster map into the raster calculator (so just duplicating it).

Does anyone have an explanation for this? I appreciate it, thank you!

Re: Raster cell count changes

DanPatterson_Retired — Fri, 22 Mar 2019 11:51:45 GMT

can you provide an example where the cell sizes and and extent for inputs and outputs are the same. That factor is just too convenient to be a coincidence for all raster tools in arctoolbox

Also software and version and extensions if any would help narrow down those issues

Re: Raster cell count changes

Roosvan_der_Deijl — Fri, 22 Mar 2019 12:11:43 GMT

Hi Dan, thanks for the fast reply!

Software = ArcMap Desktop 10.6, version 10.6.0.8321, and I'm using the spatial analyst extension.

An example would be the raster 'KK_1990_grid_cellsize60', which layer info looks like:

and top op the attribute table:

When I insert this map into raster calculator, it makes the map "kk_1990_gri1", with layer info

And raster attribute table

where you can see that all count values are 4 times smaller.

Let me know if you'd need more info to understand! Thanks!

Re: Raster cell count changes

DanPatterson_Retired — Fri, 22 Mar 2019 15:00:37 GMT

On the first image … 'Extent' tab... that is what I wanted you to check.

You have confirmed that the rows and columns of the inputs are the same. So the Extent and Spatial reference would be useful.

I also notice Dropbox. can you save your rasters to a local folder in *.tif format... just trying to rule out location and file format as well

Re: Raster cell count changes

Roosvan_der_Deijl — Fri, 22 Mar 2019 15:48:40 GMT

Extent of original map:

Extent of calculated map:

So it's the same..

When I export the file to a local folder and open it in ArcMap, it AGAIN has counts /4. So really when I do anything to the file it divides all cells counts by 4. I'm so confused now.

PS. It happens with other files as well, so it's not specific to this one.

PPS. I also tried the 'build raster attirubute table' tool to check if that fixes it: no.

Thanks again!!

Re: Raster cell count changes

DanPatterson_Retired — Fri, 22 Mar 2019 15:53:57 GMT

So if you converted using Value to Value then I haven't seen this before at all. We are either missing the obvious or there is some other issue since you have ruled out coordinate system, extent and raster cell size.... time for tech support

Re: Raster cell count changes

Roosvan_der_Deijl — Fri, 22 Mar 2019 16:00:40 GMT

Alright, thanks anyway for thinking along!

Re: Raster cell count changes

DanPatterson_Retired — Fri, 22 Mar 2019 16:16:08 GMT

only one other close... hmmm

https://community.esri.com/message/71868

Re: Raster cell count changes

Roosvan_der_Deijl — Fri, 22 Mar 2019 16:39:14 GMT

That is actually true for me too:

- my columns & rows are 2335 x 987 = 2,304,645

for both the original and the calculated maps,

- but my total cell countse (as found in classify in symbology) are

- 3,879,242 for the original map and

- 969,802 for the calculated map.

So for both that doesn't make any sense, right?

Re: Raster cell count changes

DanPatterson_Retired — Fri, 22 Mar 2019 16:49:19 GMT

skip symbology as being the go to the summary of info. That is a separate issue... I just want to make sure that there is no difference in the actual raster cell count for a 1 to 1 conversion. That is symbology thing, which may have a source of error elsewhere, which I wasn't even looking for. I have found count errors for example in charts which haven't been resolved as yet since they are down on the 'issue' chain I suppose

Re: Raster cell count changes

Roosvan_der_Deijl — Sun, 24 Mar 2019 12:14:16 GMT

Ah okay I think I understand, but how could I know whether there is a difference in the actual raster cell count?

Re: Raster cell count changes

Anonymous User — Mon, 25 Mar 2019 14:37:07 GMT

Hi Roos,

The easiest way to access the actual raster cell counts is to right-click the raster in the Catalog window and select "Properties":

http://desktop.arcgis.com/en/arcmap/10.6/manage-data/raster-and-images/raster-dataset-properties.htm

I agree with Dan at this point, it might be time to contact Tech Support if the actual raster properties are the same too.

Happy rastering!

-Lauren