Select to view content in your preferred language

Repair Geometry Failed

1004
12
Jump to solution
03-14-2025 06:50 PM
SydneyHeathcott
Occasional Contributor

When I trying to use the Feature to Raster tool I get these horizontal streak lines, another post said that the Repair Geometry tool should help to fix this and when trying to run the Repair Geometry  tool I get this error message

999999: Something unexpected caused the tool to fail. Contact Esri Technical Support (http://esriurl.com/support) to Report a Bug, and refer to the error help for potential solutions or workarounds.

I tried pulling the feature into a blank project to run in case it was an issue with the project and the same thing happens. I then thought I'd try to go a round-about way by doing Feature to Polygon then Polygon to Raster, but Feature to Polygon failed with this error message

160196: Invalid Topology

0 Kudos
1 Solution

Accepted Solutions
SydneyHeathcott
Occasional Contributor

this is an update - I downgraded back to 3.3 and started my project over and it has fixed this issue.

View solution in original post

12 Replies
DanPatterson
MVP Esteemed Contributor

Repair geometry works on vector geometry, to ensure that is what you are trying to do first. 

Repair geometry does not work on raster data.

If what you are trying to "repair" is indeed vector-based, what type of data is it?  A featureclass in a file geodatabase?  So web-based thing?  Cad data?  etc etc


... sort of retired...
0 Kudos
SydneyHeathcott
Occasional Contributor

@DanPatterson

im trying to repair a feature class that is in a geodatabase, its originally ssurgo data that i have edited down to a single sand type. I've restarted this project several times because at several points ive been unable to find the causes of errors, however, ive just duplicated the errors. This is the most trouble ive had with a project so ive included some screenshots of my notes to give a better understanding of everything ive done - they are unedited sorry

0 Kudos
DanPatterson
MVP Esteemed Contributor

well at least you have the brilliant guy with the extract by mask idea going for you. 

Now back to the geometry. 

The original featureclass should be check for geometry errors, that is the one first used in the project.  No clipping, messing with borders etc.

....ssurgo data that i have edited down to a single sand type....

if you are only interested in one sand type, then do a Select By Attributes to get all the records for that type, then Copy Features to get a "clean" (hopefully) featureclass of just that sand type.  No editing geometry

Run a Check Geometry on that file.  If it needs to be repaired, run Repair Geometry.

once you get a clean file with the one sand type, clip it so size or whatever. and proceed from there.

I would review your notes if that doesn't give you a starting point.


... sort of retired...
0 Kudos
DavidPike
MVP Notable Contributor

Agree with Dan.  I'd export to a new Feature Class first (after selecting SDand only as Dan said).  That may exclude the troublesome features from the export, so I would also then do a select by location (input=originalFeatures, selectFeatures=newFeatures) and see if any features from the original are missing from the selection.

Then check Geometry - not repair - it may give you more details rather than a blanket error.

0 Kudos
SydneyHeathcott
Occasional Contributor

@DavidPike  @DanPatterson 

Thank you for the suggestions I will try this out. I've ran the Check Geometry tool several times throughout all this and often get these warning errors:

WARNING 000442: short segments at 3554 in ne_sands3

WARNING 000442: short segments at 1734 in midwest_sands_test

which was why I was running the Repair Geometry tool. I had previously ran the Check Geometry tool on the original but it came back with no errors so that lead me to believe that it was something I was doing after that.

I started it again around 11am today and it's still running but when I hover over the progress bar it shows this warning message

WARNING 000442: could not find spatial index at -1 in Map Unit Polygons - CONUS

the properties show that there is a spatial index attached to the feature. Hopefully I'll have an output soon. 

0 Kudos
RhettZufelt
MVP Notable Contributor

Are you using enterprise geodatabase or File geodatabase?

I have ran into issues trying to repair geometry with SDE data.  Even if it says it completed, copy to FGDB and the errors are still there.

Don't know if this is related to your issues, but can't hurt to check.

R_

0 Kudos
SydneyHeathcott
Occasional Contributor

It's a file geodatabase feature class. it's still trying to check geometry since 11am, I was really hoping I could get away with using 3.4 but I'm starting to consider uninstalling and going back to the older version to see if it will run there.

0 Kudos
SydneyHeathcott
Occasional Contributor

this is an update - i started a new project because it was taking too long to run and the fresh data had some warnings with short segments after running the check geometry tool. I tried to then run the repair geometry and got the same 999999 error message. 

0 Kudos
DanPatterson
MVP Esteemed Contributor

Sort the ShapeLength field in ascending order.

What appears?  Select one or two of the rows and Zoom To Selected?

Are they real?  Are they needed?  Can you delete them?

It might solve trying to repair something that isn't worth it


... sort of retired...