How do I use python to batch process the reprojection of a group of shapefiles?

Question

My client would like his shapefiles reprojected and I would like to get a script that can batch process all the files located in the same folder. I am new to Python and ArcPy but am very interested. We are using ArcMap10 as our platform software. Any help or advice would be greatly appreciated.

Answer 1

The batch project tool is designed for just such a task

The help files provide a really clear standalone python example, which can be adapted to meet your needs:

# Name: BatchProject.py
# Description: Changes coordinate systems of several datasets in a batch.

import arcpy
from arcpy import env

# Set workspace environment
env.workspace = "C:/data/wgs1972.gdb"

# Input feature classes
input_features = ["cities", "counties", "blocks", "crime"]

# Output workspace
out_workspace = "C:/data/output.gdb"

# Output coordinate system - leave it empty
out_cs = ''

# Template dataset - it has GCS_WGS_1984 coordinate system
template = "C:/data/wgs1984.gdb/stateparks"

# Geographic transformation - 
transformation = "WGS_1972_To_WGS_1984_1"

try:
   res = arcpy.BatchProject(input_features, out_workspace, out_cs, template, transformation)
   if res.maxSeverity == 0:
      print "projection of all datasets successful"
   else:
      print "failed to project one or more datasets"
except:
   print res.getMessages()

And here is my slight modifications of the script to run on a collection of shapefiles of unknown names, with my additional comments in block capitals, to try and outline what is happening where.

# Name: BatchProject.py
# Description: Changes coordinate systems of several datasets in a batch.

import arcpy, os #USE OPERATING SYSTEM FUNCTIONS
from arcpy import env

#THIS IS THE PATH TO YOUR DATA
mypath = "C:/data/"

#THIS IS AN EMPTY LIST WHERE WE WILL STORE OUR SHAPEFILE NAMES
shapefiles = []

# Set workspace environment
env.workspace = mypath

# Input feature classes
#HERE WE WILL USE THE OS.PATH FUNCTIONS TO GET A CRUDE LIST OF ALL
#OF THE SHAPEFILES IN THE DIRECTORY, BY GETTING A LIST OF ALL FILES
#IN THE DIRECTORY AND THEN FILTERING OUT ONLY FILES WITH THE .SHP EXTENSION
for filename in os.listdir(mypath):
    if filename.endswith('.shp'):
        shapefiles.append(filename)

# Output workspace
#THIS IS WHERE THE REPROJECTED SHAPEFILES WILL BE SAVED, THIS FOLDER MUST EXIST
out_workspace = "C:/data/output" 

# Output coordinate system - leave it empty
#IF YOU ARE NOT USING A TEMPLATE SHAPEFILE,
#YOUR PROJECTION PARAMETERS WOULD GO IN HERE
out_cs = ''

# Template dataset - it has GCS_WGS_1984 coordinate system
template = "C:/data/stateparks.shp"

# Geographic transformation - 
transformation = "WGS_1972_To_WGS_1984_1"

#THE TRY/EXCEPT STRUCTURE IS USED TO CATCH ANY ERRORS AND RETURN THE ERROR MESSAGES
try:
   #THIS LINE SETS UP THE BATCHPROJECT PARAMETERS
   res = arcpy.BatchProject(input_features, out_workspace, out_cs, template, transformation)
   if res.maxSeverity == 0: #THIS RUNS THE PROJECTION FOR ALL OF THE LISTED LAYERS
      print "projection of all datasets successful"
   else:
      print "failed to project one or more datasets"
except:
   print res.getMessages() #DISPLAYS ANY ERROR MESSAGES

This code is quite extensive; the same result could be achieved in many fewer lines, however this (I hope!) gives you a good overview of the arcpy object.

Answer 2

If you want your output as a shapefile instead of a geodatabase, you can use the second codeblock of sgrieve and replace the try: part with the following:

try:
   for shapefile in shapefiles:
       arcpy.Project_management(mypath + os.sep + shapefile, mypath + os.sep + "porj" + shapefile, out_cs, transformation)
   else:
      print "failed to project one or more datasets"

The output shapefiles will be renamed with proj as prefix.