Why does ArcMap edit session show different vertex count on feature than arcpy/ ArcToolbox

Question

I am looking at features in an ArcSDE geodatabase via ArcMap. The feature shows to have four vertices by viewing the vertices in the map and by looking at the sketch properties.

However if I run a feature vertex to point tool i get 5 vertices. If I get an arcpy.SearchCursor row object and do row.pointCount I get 5 vertices reported.

I even exported the feature class out to a shapefile and got the same results.

Whats going on here? Thanks.

• najuste: Isn't the feature a polygon? Then if the polygon ring is closed, you have the first and the last point the same..
• Chad Cooper: Are these polygons? If so, you may be seeing the start and end point (which are the same) showing up twice. Not in front of ArcMap right now to test, but will in a bit - @najuste beat me to the comment by a few seconds...

Answer 1

It sounds like you are working with polygons. I created a 4-corner polygon in a FGDB, and when I look at it in the editor, it does have 4 vertices:

enter image description here

However, when I run the following script on it, you can see it does indeed have 5 vertices (as it should), where the first and last are the same in order to close the polygon:

import arcpy
infc = "NumberOfVertices"
desc = arcpy.Describe(infc)
shapefieldname = desc.ShapeFieldName
rows = arcpy.SearchCursor(infc)
for row in rows:
 feat = row.getValue(shapefieldname)
 print "\tFeature %i:" % row.getValue(desc.OIDFieldName)
 partnum = 0
 for part in feat:
 parts = []
 print "Part %i:" % partnum
 for pnt in feat.getPart(partnum):
 if pnt:
 parts.append([pnt.X, pnt.Y])
 else:
 parts.append(" ")
 partnum += 1 
 print parts 
Part 0:
[[414609.77850000001, 3711987.8151999991], [421594.30240000039, 3711987.8151999991], [421926.89879999962, 3700346.9419999998], [414277.18209999986, 3701178.4329000004], [414609.77850000001, 3711987.8151999991]]
>>> len(parts)
5
>>>

• blah238: Yeah, my guess is the last vertex is hidden to protect the user from accidentally making the geometry invalid.
• Chad Cooper: @blah238 Very good point.