I have got a lot of point shapefiles (r20010101.shp, r20010202.shp, r20020101.shp...). Each of them has the same amount of points at exactly the same location. In a column that is named after the Date e.g. "D20010101", where I have got unique values for every shapefile.

What I want to do now, is to Combine or Merge or Join all these shapefiles in a Python command.

The result should be ONE shapefile with all those point (but please NO duplicate ones!!).

The attribute table of this shapefile should look like this: FID, Shape, Id, [Value of Date 2001/01/01], [Value of Date 2001/02/02],...

  • Sorry i forgot, I am Using ArcGis 9.3.1!
    – Markus
    Jun 15 '11 at 12:48
  • I tried fiddling with the idea using OGR, I converted everything into spatial lite I think then back as one shp file. I'll try to dig up some code
    – dassouki
    Jun 15 '11 at 14:53
  • One question\comment: unless the original point-shapefiles contain exact duplicates within themselves (that is: r20010101.shp has 2 points in the same place with exact same info), How are you going to decide which duplicate to erase? As you pointed out, although point1 from r20010101.shp and point2 from r20010202.shp are in the same coordinates, they are NOT full duplicates as they contain different data...
    – jonatr
    Jun 19 '11 at 9:25

NOTE: This works in ArcMap10. You will have to test it in 9.3. -- Join the three shapefiles together using the Join tool. Make sure you have an indexed feature that can be used to join them on (it speeds the tool up dramatically). After doing that, you can use the Fields tab inside the shapefile / layer properties to make certain fields visible/hidden. If you hide fields and export the layer to a new feature class / shapefile, the resulting file will only have the visible fields.

Alternatively, once you have the join, you can export it wholesale and perform any number of tools upon it:

  1. Use merge and mangle the field map (allows renaming of fields). Yes, you can perform a merge on a single layer just to change names and whatnot.
  2. Use ModelBuilder to create a batch Field Delete python script for your shapefiles.
  3. Add Field to get new names and calclulate the fields you want with the old data. Then you can delete the old fields with their ugly qualified names.

Last, and least (slow process is slow), you could create new fields and cursor through your data, reading the date information from the other tables row by row.

So, as you can see, many options! And all of it is accessible through the GP processor (although I find joining multiple layers at once to be unreliable in a script).


This is what I was looking for.

import arcgisscripting, os
import sys

gp = arcgisscripting.create(9.3)
gp.Workspace = "D:\\Workspace\\shapedata"
# With Files:
# r20010101.shp with Fields and Input like:
# FID,Shape,POINTID,D20010101
# 0, Shape, 1, 10
# 1, Shape, 2, 14

# r20020101.shp with Fields and Input like:
# FID,Shape,POINTID,D20020101
# 0, Shape, 1, 12
# 1, Shape, 2, 13

# r20020601.shp with Fields and Input like:
# FID,Shape,POINTID,D20020601
# 0, Shape, 1, 11
# 1, Shape, 2, 9

# r20030101.shp with Fields and Input like:
# FID,Shape,POINTID,D20030101
# 0, Shape, 1, 3
# 1, Shape, 2, 7

fcs = gp.ListFeatureClasses("", "POINT")

for fc in fcs:
    outfc = "D:\\Workspace\\shapeout"
if x==0:
    # just run once:
    fc_old=fc #r20010101.shp
    fn_old="D"+fc[1:-4] #D20010101
    # do this, as long as there are shapefiles in gp.workspace
    # examples given here for the FIRST SpatialJoin

    fc_new=fc # r20020101.shp
    outfc=outfc+str(x) # D:\\Workspace\\shapeout1
    fn_new="D"+fc[1:-4] #  D20020101

    fieldmappings = gp.CreateObject("FieldMappings")
    vt = gp.CreateObject("ValueTable")


    if x>1:
        # This should prevent Fields "Join_Count", "Join_Co1", "Join_Co2"....

    gp.SpatialJoin(fc_old, fc_new, outfc, "#", "#", fieldmappings) 

    if x>1:
        # delete temporary file e.g. shapeout1.shp after next loop 

    fc_old=outfc + ".shp" # D:\\Workspace\\shapeout1.shp will be joined in next loop with e.g. r20020601.shp


# The final Output will be 1 Shapefile with Fields and Input like: 
# FID,Shape,POINTID,D20010101,D20020101,D20020601,D20030101
# 0, Shape, 1, 10, 12, 11, 3
# 1, Shape, 2, 14, 13, 9, 7

As I have got 30 Pointfiles (150MB each), this takes pretty long on my Computer (not too old), like 1-2 hour for every SpatialJoin!

  • 2
    1-2 hours, yikes! Did you have spatial indexes for each file?
    – Mike T
    Jul 6 '11 at 10:56

if the "ID" field is unique within each table and consistent for all point across the many shapefiles, you can just use the join field tool under Data Management - Joins - Join Field.

as Nathanus stated, make sure the ID field of each shapefile is indexed.

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.