topic Re: Focal statistics problem values in ArcGIS Spatial Analyst Questions

Focal statistics problem values

RobertoFilloramo — Thu, 20 Jun 2019 15:01:32 GMT

Hi,

I am using ArcGIS 10.1 and I am working with the Focal Statistics tool in order to get a smoother surface (I set "neighborhood rectangle 3x3 cell units" and as statistics type "mean"). The obtained result it looks like not so fine since the focal st. values seem increased respect to the input raster (see figures), but I don't understand the reason. The doubt arose making some comparison with other results in which the focal st. raster values are smaller than the input raster.

Could someone help me?

Thank you

Re: Focal statistics problem values

LanceCole — Fri, 21 Jun 2019 03:31:53 GMT

Roberto,

Seeing the zero (0) in your data makes me wonder if you may have holes or voids in your original data. You may want to try and combine focal statistics with another function such as Elevation Void Fill, Fill, or possibly even Nibble depending upon your data type.

Re: Focal statistics problem values

DanPatterson_Retired — Fri, 21 Jun 2019 03:57:39 GMT

from Pro but the same for Map

Focal Statistics—Help | ArcGIS Desktop

The last optional parameter can have an effect on the results.

ignore_nodata (Optional)

compare the results with this set to DATA (default) versus NODATA

Re: Focal statistics problem values

RobertoFilloramo — Fri, 21 Jun 2019 09:24:17 GMT

Thank you Lance Cole‌ and Dan Patterson‌ for your response. As suggested I tried to compare the focal st. results ignoring the NODATA, but in both cases the values were bigger than the input raster (see fig. "FocalSt_IgnorNoData" stands for the Ignore NoData flagged option, contrary to "FocalSt_NoData").

For what concerns the holes/voids, I used the Fill tool. The output raster had bigger values (see fig.), I presumed because of the filling operation, but using the new output for the focal st. I noticed the same problem (see fig.).

I am just wondering now if the problem is since the raster file "ALS_Erwitte_geotif_05m.tif" was generated in CouldCompare using nearly 8 get-together tiles. Each tile was represented by a .csv containing the coordinates points of the entire area.

What do you think?

Thank you!

Re: Focal statistics problem values

DanPatterson_Retired — Fri, 21 Jun 2019 10:41:13 GMT

What is the nodata value in the input rasters set to in all cases?

And can you confirm that you have nodata areas in the rasters especially the ...05.tif file. That one may be the culprit

Re: Focal statistics problem values

RobertoFilloramo — Fri, 21 Jun 2019 11:41:40 GMT

Concerning the "nodata value in the input rasters set", I meant the "ignore_nodata" set (optional) in the focal st. window tool.

Regarding the "...05.tif" file, it showed no nodata, as you can see in the fig.

Therefore, according to what you wrote the nodata lack in the input file ("...05.tif") may be the problem, at this point. In order to get a smoother surface, how could I solve this issue?

I wanted a smoother surface to use it as input for the least cost path analysis, in order to avoid that the process could take into account small and medium surface variations/modern structures (such as a street or a roundabout) during the path estimation.

Re: Focal statistics problem values

DanPatterson_Retired — Fri, 21 Jun 2019 13:58:23 GMT

Roberto, the nodata value may not have been set, it doesn't mean that there are areas that shouldn't be treated as nodata.

If this is a dem, it would be unlikely that you would have '0' unless your area is at sea level. If it is some other file type (ie landuse), then It would be useful to know what the areas of low value (near or at 0) represent

Re: Focal statistics problem values

RobertoFilloramo — Fri, 21 Jun 2019 15:20:38 GMT

Well, the file is a DGM and the csv file (forming the tiles) containing just x, y and z information. Unfortunately, the site from which I downloaded the files didn't say anything about nodata and the only data, about min and max values that I have, are those given by the raster value scale in the layer panel (as in the previous fig.).

That's the problem.

Re: Focal statistics problem values

DanPatterson_Retired — Fri, 21 Jun 2019 17:31:52 GMT

If the lowest elevations ie 0's don't make sense for your area (ie sea level) there is a good chance that 0 is the nodata value.

It would be worthwhile looking at the histogram of the symbology to see it that is indeed the case

Re: Focal statistics problem values

RobertoFilloramo — Fri, 21 Jun 2019 18:22:07 GMT

I am afraid that, as you said, the 0 represents the nodata. Also in the histogram is reported the lowest value as 0 (see fig.). So the data has been stored like this, although not being at the sea level (it is an area of northern Germany).

Re: Focal statistics problem values

LanceCole — Fri, 21 Jun 2019 20:12:08 GMT

Dan, I agree. Typically seeing zero (0) data is a red flag something is up unless expected.

Re: Focal statistics problem values

DanPatterson_Retired — Fri, 21 Jun 2019 21:18:40 GMT

Time to set the raster nodata is set so it doesn't foul up the focal stats in the areas where nodata borders data

Re: Focal statistics problem values

JuanLaguna — Sat, 22 Jun 2019 01:12:36 GMT

Hi Roberto.

I'm sorry you're having some difficulty. Let's see if we can help get this sorted out.

First off, I notice the your input raster is identified as a 64-bit double-precision TIFF. Is this intentional? I don't often come across this particular type of Tiff data. If you use the Copy Raster GP tool to make a new version of the data with the Pixel Type set to a regular 32-bit float, does the resulting raster have the same outcome when used in the Focal Statistics tool?

If you convert your input raster to an Esri GRID format raster and use that as input, does the problem recur?

I did some testing with some local data that I mapped to a value range similar to what you report, but could not produce any unexpected results. Would it be possible to provide a copy of your data to test with? If you could prepare a smaller section of it that still demonstrates the problem, then it would facilitate the investigation.

Thanks,

Juan

Re: Focal statistics problem values

RobertoFilloramo — Sat, 22 Jun 2019 13:48:46 GMT

thank you for your suggestion, I uploaded the sample tiff and a csv sample tile in order to have an idea on my data. The tiff is originated from the csv in CloudCompare software.
Moreover, I tried the tools that you suggested but the result is the same (see fig.)

Re: Focal statistics problem values

JuanLaguna — Sat, 29 Jun 2019 02:41:00 GMT

Hi Roberto.

As others had suggested, the issue does seem to be a result of having cells with a value of 0 (instead of NoData) present in the input.

Highlighted below is a section of the input demonstrating some of those cells:

If you were to perform a Con operation on the input to convert those 0 values to actual NoData:

The final result should be more what you were expecting:

Hope this helps!

Cheers,

Juan

Spatial Analyst team

Esri, Inc