I have written many Python scripts using ArcPy in ArcGIS 10, and so far my only means of debugging is restricted to printing messages to the geoprocessing results window using arcpy.AddMessage().

Are there any other options out there, such as setting break points?

Jason's method works great. If you have a bug in your toolbox, such as validation, your IDE probably won't be able to pinpoint the problem because toolboxes are encoded. At least WING wasn't able to pinpoint it.

  • Maybe an option here? code google pyscripter – Brad Nesom Mar 2 '11 at 19:55
  • The ArcGIS help discusses how to use the debugger in PythonWin – Regan Sarwas Mar 3 '11 at 1:17
  • In the OS file properties, have you tried setting the default program for .py files to PyScripter? – KevinMayall Apr 7 '11 at 19:18
  • By doing this, wouldn't I be effectively telling the OS that I didn't want to run my python script, but rather just open it in the editor? – RyanDalton Apr 7 '11 at 20:05
  • Yes, correct. Sorry I misunderstood your purpose. – KevinMayall Apr 11 '11 at 17:38

Usually Python debuggers/IDEs assume the Python script is running in the same process as itself so debugging a script running in ArcMap.exe is right out -- you need to get enough of the GP scripting environment bootstrapped in a Python script as you can to debug with.

A method that's worked very well for me over the past few years is to write a simple script that just calls the tool and use that as my main script in the Python IDE (Wing or Pythonwin) and have my breakpoints set in the tool's .py file also open in the same IDE session.

So basically I do this:

  1. Get the set of inputs that aren't working in my script tool
  2. Open a simple .py file in the same folder as the .tbx that calls the tool
  3. Open up the caller script and the script tool .py file in the IDE
  4. Set breakpoints in script tool file
  5. Run the caller script

And my caller script is usually pretty simple:

import os
import arcpy
arcpy.ImportToolbox(os.path.join(os.path.dirname(__file__), 'my.tbx'))
arcpy.MyToolThatIsFailing_myalias("inputs", "that", "don't" "work")

I've tried winpdb to debug scripts running in ArcMap but I've never had any luck. If you want to try it out and you get it working well, please share your findings.

  • 1
    nice solution! i will try it. – George Silva Mar 2 '11 at 20:12
  • I have used winpdb successfully in the past to debug long-process scripts. winpdb, once set up allows you to "jump into" a currently running standalone script or theoretically one within another process such as ArcMap and debug it normally. This was very handy for checking in on a script that would take weeks to complete. I have not tried debugging arcpy within an ArcGIS application process though. – blah238 Mar 3 '11 at 4:47
  • I'd recommend PyCharm. it works like magic – user39901230 Apr 4 '11 at 21:26

You can change it from the GP options dialog, just point to your executable of choice for editor/debugger.

GP Settings Dialog

  • This looks to be a new enhancement for ArcGIS 10, correct? I don't see this setting option for 9.3. – RyanDalton Apr 7 '11 at 20:25
  • Correct, new in 10. – Jason Scheirer Apr 7 '11 at 20:33

For users of 10, clearly the best route to go is with Jason's post, which I marked as the "best answer". For users of 9.3, I was able to follow the ESRI KB Support instructions linked in Brad's post to get it working.

Ultimately the key to get to default editing of ArcGIS python scripts in Pyscripter was to edit the system "action" for "registered file types" ending in *.py files (step #4). I created a new "Edit" action type and then included the Pyscripter.exe path. Once I did this, the default edit action was setup to launch Pyscripter instead of IDLE.

The string I used (because it was cut off in the dialog box shown below) is:

"C:\Program Files\PyScripter\PyScripter.exe" "%1"

enter image description here

  • 1
    +1 for the answer, and again, if I could, for including the string and not just an image (text is so much easier to copy and paste!). – matt wilkie Sep 20 '11 at 22:39

I don't know how this will work with arcpy, but you can try pdb:

import pdb; pdb.set_trace()
  • there is a similar functionality for pydev with eclipse, which is described here. You set eclipse to listen for a signal, and you can trigger that in code running entirely outside eclipse by a pydevd.settrace() function call. urbansim.org/Documentation/DebuggingWithEclipse – ako Jul 29 '11 at 6:14

The most straight forward way I have found to debug my code is to hard code my inputs to an existing dataset's file path. For example:

import arcpy

# parameters for script
#input1 = arcpy.GetParameterAsText(0) # layer
#input2 = arcpy.GetParameterAsText(1) # folder

# debugging parameters
input1 = r"path/to/your/data.shp"
input2 = r"path/to/your/folder"

In addition to the results window, the python window yields helpful error messages and/or allows you to test code snippets.

Any good IDE has a debug probe that will have, in memory, all your processing up to the current breakpoint. With that, you can see what is going on with the data. Set breakpoints where you want to pause the script. Use conditional breakpoints if you want to stop a loop at a specific iteration. Also, know your Arc versions and what functions are available to those versions.

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.