0

I am trying to create and update a field with a count of line features (tLayer) within a distance of point features (sLayer). I am attempting to use a combination of AddField_management, arcpy.da.SearchCursor, SelectLayerByLocation_management, arcpy.GetCount_management, and arcpy.da.UpdateCursor.

The code I have for this is currently updating all records for the Line_Count field with the count of the point features (i.e. 2) for only the last record. Though, a print statement following the GetCount line will return the line count for all of the point features (with a few unessential iterations). I believe it is only printing the last row in 'result' for the Line_Count update statement. Is there an efficient way for result (i.e. result=int(arcpy.GetCount_management("tLayer_lyr").getOutput(0))) to be saved as a list or an array that can be written to the Line_Count field? ...Or would it be best to use a dictionary?

What do I need to do to appropriately update the Line_Count field for all of the records? Also, this process will be applied to a large dataset and will be extended to include 'where clauses'. Are there any suggestions as to how to make this as efficient as possible?

import arcpy
from arcpy import env
arcpy.env.OverwriteOutput = True

defaultGdbPath = 'C:\Topo_Check_Tess_V5.gdb'

sLayer='C:\Topo_Check_Tess_V5.gdb\Stations'
tLayer='C:\Topo_Check_Tess_V5.gdb\Lines'
#ppLayer='C:\Topo_Check_Tess_V5.gdb\Plants'

arcpy.AddField_management(sLayer, "Line_Count", "SHORT", "", "", "", "", "NULLABLE", "NON_REQUIRED", "")

TLineCountField = "Line_Count"


arcpy.MakeFeatureLayer_management(tLayer, "tLayer_lyr")
with arcpy.da.UpdateCursor (sLayer, TLineCountField) as LineCountList:
       for s_row in LineCountList:
            with arcpy.da.SearchCursor(sLayer,["OBJECTID", "SHAPE@"]) as s_list:
                    for station in s_list:
                        arcpy.SelectLayerByLocation_management("tLayer_lyr", 'WITHIN_A_DISTANCE', station[1],.000002, "NEW_SELECTION")
                        result=int(arcpy.GetCount_management("tLayer_lyr").getOutput(0))
                        print result
                        LineCountList.updateRow(s_row)
                    del station 
            del s_row

print "Done"

Updated Line_Count Field Print readout

7
  • Add s_row[0] = result before LineCountList.updateRow(s_row) – klewis Feb 14 '17 at 23:51
  • You might be able to use the Generate Near Table tool to accomplish the same thing. – Bjorn Feb 14 '17 at 23:52
  • for each row in the update cursor, it is running through every row in the search cursor, so you're only ever getting the result from of the final row of the search cursor. make sense? – Adam Feb 14 '17 at 23:59
  • Why the needless scripting? Create a frequency table, then join and calculate pro.arcgis.com/en/pro-app/tool-reference/analysis/frequency.htm – Ben S Nadler Feb 15 '17 at 0:13
  • 1
    Nesting cursors and expensive GP calls for spatial operations is terrible for performance. Unravel all those for loops and see where it can be optimized. – Paul Feb 15 '17 at 1:17
1

It is generally not an efficient approach (as several have commented) to use cursor processing in this way, but if you accept the performance hit (maybe you're not working with large feature classes), then you should at least eliminate the nested cursor. The update cursor can also be used as your 'read' cursor, so something like the following rewrite of your outer with block should work - this is untested:

with arcpy.da.UpdateCursor (sLayer, [TLineCountField, "SHAPE@"]) as LineCountList:
    for s_row in LineCountList:
        arcpy.SelectLayerByLocation_management("tLayer_lyr", 'WITHIN_A_DISTANCE', s_row[1], .000002, "NEW_SELECTION")
        result=int(arcpy.GetCount_management("tLayer_lyr").getOutput(0))
        print result
        s_row[0] = result
        LineCountList.updateRow(s_row)
1
  • Thank you soooo much. This works perfectly! I will be working with large feature classes, so I will need to look into ways to optimize efficiency. I'll look into some of the suggestions for this. Thanks again! – Tess Gardner Feb 15 '17 at 15:40

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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