Error while working with zonal statistics as table

[LauraSochas]

Hi everyone!

Thanks for reading this post -

I'm trying to get "zonal statistics as table to work", but am encountering an error "background processing has encountered a serious error and is unable to continue". The error log code (in case that is relevant) is "000000_12301899".

After I switched off background processing, the execution started but stalled at 91%.

My zone dataset is an integer raster with about 1200 zones, and my zone field is a string. My value dataset is a raster with values between 0 and 48.7. I am only interested in the SUM function.

My zone and value datasets have 17,302 columns and 10,937 - if you think this is too much for ArcGIS to process, could you recommend practical ways to segment the data such that my zones are not cut up?

Any help would be greatly appreciated - happy to give more details!

Thanks in advance

Laura

[DuncanHornby]

Here are some ideas:

• Are your 2 rasters the same coordinate system? If not re-project one so it matches the other.


• Do the cells overlay each other, i.e. you are not seeing some sort of misalignment?


• Are they the same cell size, if not re-sample the zonal layer to match the data layer


• Your number of columns suggest that they have different extents so your zonal layer may not cover your data layer, not sure that is a problem but try to fix that?

Duncan

[NidhinKarthikeyan]

Laura,

The forum post by William Huber might help you.

[LauraSochas]

Thanks very much Duncan and nidhinkn! Much appreciated!

I checked what you recommended, which was very useful, however was still getting errors. I turns out that the problem was that much of my zone data had "no data" in places where the value dataset did have data. I converted all "no data" to 9999, using Con and IsNull in the raster calculator and this time it worked.

Many thanks for your help!

Laura

[curtvprice]

Laura, I don't see why this should have broken the tool.

Esri developers can make great use of datasets that break their tools, so I strongly recommend you submit a support incident so they can figure out why your workaround was necessary (and maybe fix a problem before the rest of us run into it!)