I have a destination (Feature Class):


I have a database table in the same file geodatabase that contains my source point data. I have columns with attributes in addition to the requisite coordinate data.

Using Arcpy, how do I retrieve the points from the source database table and create points in the destination Feature Class?

I tried creating a single point using this code:

import arcpy
cur = arcpy.InsertCursor(fc, ["SHAPE@XY"])
pnt = arcpy.Point()
pnt.x = 20000
pnt.y = 10000
feat = cur.newRow()
feat.shape = pnt
del cur

but, although a row is created, the coordinates are 0,0.

Current thoughts - I'll probably use a cursor to step through the records retrieved and for each row in the source, build a point and insert it into the destination. The logic seems to make sense.

3 Answers 3


I am not sure whether nested cursors (e.g. a SearchCursor for the source table and an InsertCursor for the destination feature class) will work, but it's worth a shot.

Specific to your problem of how to get geometry data into a feature class using an InsertCursor:

  1. Use a da.InsertCursor (since you have access to 10.1+)

  2. Take a look at the Help page for arcpy.da.InsertCursor, particularly the examples. I think that you are making it more complex than necessary.

This is their example. No geometry objects needed, just the numeric values.

import arcpy

# A list of values that will be used to construct new rows
row_values = [('Anderson', (1409934.4442000017, 1076766.8192000017)),
              ('Andrews', (752000.2489000037, 1128929.8114))]

# Open an InsertCursor
cursor = arcpy.da.InsertCursor("C:/data/texas.gdb/counties", ("NAME", "SHAPE@XY"))

# Insert new rows that include the county name and a x,y coordinate
#  pair that represents the county center
for row in row_values:
  • I could not get the two nested cursor approach to work. I tried but I'll admit I didn't spend much time trying to figure it out. In the end, I used a SearchCursor to populate an array and then an InsertCursor to insert the items from the array. It works and for 20K records, it finishes in < 3 seconds. Commented Feb 4, 2015 at 14:50
  • An array for storing data is my usual go-to method, but it can be slightly cumbersome with many attributes to transfer. Glad you got it to work :)
    – Erica
    Commented Feb 4, 2015 at 16:04

I think you would be better off using the Make XY Event Layer and then use Copy Features to export it to a new feature class.

import arcpy

def xy_table_to_fc(table, x, y, wkid, out_fc):
    lyr = str(arcpy.management.MakeXYEventLayer(table, x, y, 'temp', wkid).getOutput(0))
    arcpy.management.CopyFeatures(lyr, out_fc)

if __name__ == '__main__':
    table = r'C:\TEMP\Headstones_table.dbf'
    out = r'C:\TEMP\Headstone_points.shp'
    xy_table_to_fc(table, 'long', 'lat', 103793, out)

You're just mixing up your cursors a bit. Some of your code is from the old cursor type. Some is from the new data analysis type.

"SHAPE@XY" isn't supported in the old version of an InsertCursor.

Use da.InsertCursor instead of the old version and...

Instead of:

feat.shape = pnt


feat[0] = (20000, 10000)

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.