I have two shapefiles. Shapefile A has five attributes and shapefile B has 20 attributes. I use the Identity tool to determine, what of shapefile A lies in shapefile B.

This is how both attribute tables look like:

Shapefiel A


Shapefile B


My results are rather confusing


How comes that it splits BO, FE, ... up, when that is contained in Reg1?

Why is it contained two times in there?

Working with ArcGis Desktop 10.2.2 here.

Here is the snipped of the python code and a screenshot.

# Replace a layer/table view name with a path to a dataset (which can be a layer file) or create the layer/table view within the script
# The following inputs are layers or table views: "clip4_diss_clipCopy", "dissolve_clipCopy"arcpy.Identity_analysis(in_features="clip4_diss_clipCopy",identity_features="dissolve_clipCopy",out_feature_class="C:/Users//Documents/Scoring/up.shp",join_attributes="NO_FID",cluster_tolerance="#",relationship="NO_RELATIONSHIPS")

The underlaying green shapefile was used as the Identity Feature. The cyan lines are the different Reg X and the dark outlined features are the Infeatures for the tool!


  • 2
    Do you now need to dissolve?
    – GISHuman
    Jul 17, 2014 at 19:26
  • Nope. I dissolved before. Thats how I came up with five types. They were all splitted up (10.000+ multipart feature). A dissolve would not work here cause it takes away information. Now I have the DISTRICT_N connected to a WLType. But there should not be two BO in Reg1 for example. That should be one!
    – four-eyes
    Jul 17, 2014 at 19:31
  • Are you trying to combine the attributes of both shapefiles into one? Have you looked into a spatial join if that is the case? Just using the Identify tool will identify those features found where the mouse clicks. It won't combine the shapefiles that overlap.
    – Branco
    Jul 17, 2014 at 19:52
  • 1
    I think the reason Identity produces the results you are observing will be very easy to explain if you include a picture of just Reg1 and anything that passes through it. Also, include the exact syntax used to run the tool which is easiest to grab by using Geoprocessing | Results to access Copy As Python Snippet.
    – PolyGeo
    Jul 17, 2014 at 21:49
  • 2
    As I explained previously, any overlay operation combines layers into areas of unique overlap, and there may be more than one such area for two shapes. The Union explanation page best demonstrates this, even though Union isn't the tool you used. We keep bringing up your process because without understanding it, we can't say "oh you want to do this - you're using the wrong tool, try this instead." That's why you're getting so many comments asking what you're trying to do, not just how.
    – Chris W
    Jul 17, 2014 at 22:31

2 Answers 2


While I cannot see your geometry, it the Identity tool is probably working as it should. Consider the following example:

I have two shapefiles, Region and Area. Area has a single shape. Region has two, both of which overlap Area, one of which does so in multiple places.

enter image description here

When I run the Identity tool using Area as input and Region as identity features, I get three resulting shapes - Area not covered by any Region, Area covered by left Region (note this is a multipart feature with three components, because those three areas of overlap derive from a single polygon), and Area covered by right Region (separate from 2 because it is derived from a different polygon).

enter image description here

If you're looking to get a single record for each type, you'll have to dissolve your Identity results on both region and type field, and allow for multipart features to be created. This would be done before any calculations, else you would lose data (this begins to get into split and merge policies, which is a whole other topic/question). Alternatively you can use the Summary Statistics tool to generate a table where each region/type has its own unique row in the table, as I mentioned in my answer to your last question.

  • I understand I guess! It seems that the Identity Tool does not seem to be the right tool then. Because I want to overlap the WLType by the DISTRICT_N and every WLType which is overlapped by the DISTRICT_N shall get their attributes (the names of the districts).
    – four-eyes
    Jul 17, 2014 at 23:26
  • @Stophface Union might be a better approach, however Identity could work with some additional steps after (and I think no matter what tool you use, it's not going to be a one-step solution). Your wltype would be the input and district the identity features. You'd just Dissolve the results as I mention in the answer.
    – Chris W
    Jul 17, 2014 at 23:46
  • yeah i need to dissolve i guess. up to now i did not know that the dissolve tool can take more than one argument!
    – four-eyes
    Jul 18, 2014 at 0:15
  • @Stophface If this answered your original question then I think you should click the green Accept button.
    – PolyGeo
    Jan 6, 2015 at 7:11

What the Identity tool is good for is taking underlying data into your feature. Think of extract value to point, but, in this case, it is vector to vector data. Your input feature is your main feature you want to add data too. Your "identity feature" will be the feature you are taking the data from. If you have a polygon of a park and want to take the information from a streams layer, you can use the park as your input and streams as your identity feature. In the output, your polygon will remain with the shapes your streams layer "carved" into it with its original attribute data. Chris W shows this well in his example.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.