How to suppresss Python Toolbox from adding output datasets to the map

Question

I wrote a Python Toolbox and would like to suppress it from adding output datasets to the active map. Below a minimal example (not the actual toolbox):

import arcpy
from typing import List
class Toolbox(object):
 def __init__(self):
 self.label = "Toolbox"
 self.alias = "Toolbox"
 self.tools = [Tool]

class Tool(object):
 def __init__(self):
 self.label = "Tool"
 self.description = "Tool Skeleton"
 def getParameterInfo(self) -> List[arcpy.Parameter]:
 output_feature_classes = arcpy.Parameter(
 name="output_feature_classes",
 displayName="Output Feature Classes",
 datatype="DEFeatureClass",
 parameterType="Derived",
 direction="Output",
 multiValue=True,
 )
 return [output_feature_classes,]
 def isLicensed(self) -> bool:
 return True # always licensed
 
 def execute(self, parameters: List[arcpy.Parameter], messages):
 
 # arcpy.env.addOutputsToMap = False
 result = do_something()
 # make available for next tool
 parameters[0].value = result.feature_classes
 return result

arcpy.env.addOutputsToMap = False does apparently only work when used directly in the Python window or from a Notebook in ArcGIS Pro.

arcpy.env.addOutputsToMap
The tool's resulting output datasets will be added to the application display. The property is only applicable when used directly in the Python window or from a Notebook in ArcGIS Pro. (https://pro.arcgis.com/en/pro-app/latest/arcpy/classes/env.htm)

How can I suppress ArcGIS Pro from adding my output to the active map if this is the only tool (or the last tool of a model) being run?

(I am aware of that if I use the tool in a model, I can configure if add or not add intermediate output to the map. I am also aware of that there is a global setting in Options.)

Note: The return value of do_something() is a namedtuple and its property feature_classes is an array of created feature classes. The code behind do_something() is completely decoupled from the Python toolbox.

• Hornbydd: Can you expand your code sample and show a real example of what the function do_something() is doing. By obfuscating your code the issue may lie within that function?
• KHibma: Do you absolutely need it to be a parameter? The last time I looked at this scenario there was nothing more that could be done than you've discovered. (Maybe things have changed). But generally, if something is a parameter, the app is going to add it to the map on execution unless the global GP setting has been turned off.
• Thomas: @KHibma, originally I didn't have a derived output parameter but if I'd like to use the result in a model, I wouldn't able to do so. If the answer to my question is that it's not possible, that's also an answer and I can stop searching for a solution. ;)
• Thomas: @Hornbydd, the return value of do_something() is a namedtuple and its property feature_classes is an array of created feature classes. The code behind do_something() is completely decoupled from the Python toolbox.
• KHibma: Well - if you're using your PYT as a sub-tool to a model, and within the model this output IS NOT a model parameter, then I'd expect that it not be added to the display. But you're right, you need the PYT to have a parameter to make it chainable in the model. (just don't put the "P" on it)

Answer 1

If your output is a derived parameter, it will by displayed accordingly to the environnement variable arcpy.env.addOutputsToMap from the active map or from an other tool/model that contains your tool script.

You cannot set this option directly in your tool script with a derived parameter.