How to parse NETWORK_LAYERS results in Python?

[LaurieMhu]

I would like to use Trace with result_types=['NETWORK_LAYERS']. I am unable to drill the layers in out_network_layer.

Here is an example of an attempt:

out_netk_layer = "out_net_lay"
stream_layer_name = 'Stream'
res = arcpy.tn.Trace(in_trace_network=t_network,
			trace_type="SHORTEST_PATH",
			starting_points=starting_points,
			shortest_path_network_attribute_name="Shape length",
			validate_consistency="VALIDATE_CONSISTENCY",
			result_types=['NETWORK_LAYERS'],
			out_network_layer=out_netk_layer)
	
if res:
	total_length = 0
	feature_class_path = '{}\\{}'.format(out_netk_layer, stream_layer_name)
    with arcpy.da.SearchCursor(feature_class_path, ["SHAPE@"]) as cursor:
        for row in cursor:
            shape = row[0]
            if shape.isSelected:
				total_length += shape.length
	arcpy.AddMessage("Total length : {} meters".format(total_length))
else:
	arcpy.AddWarning("no trace result")

arcgisscripting.ExecuteError: Execution failed. The parameters are not valid.

ERROR 000732: Input Entities: Dataset out_net_lay/stream does not exist or is not supported

Where can I find a code example? What anomaly do you see here? Thanks for everything

[Dan_Joyce_OE]

I'm encountering the same issue.  I'm expecting the trace to return a Group Layer object that could then be manipulated in the same way that the old Geometric Network trace results could be processed.  It appears the new trace only returns a string when the result type = 'NETWORK_LAYERS'.

trace_result = arcpy.tn.Trace(network, 'CONNECTED', start_points, lyr_valve, include_barriers='INCLUDE_BARRIERS', validate_consistency='VALIDATE_CONSISTENCY', result_types='NETWORK_LAYERS', out_network_layer='TRACE1')

print(trace_result)

> Trace_Results_Aggregated_Points

It's frustrating because the script works as intended within ArcGIS Pro but doesn't as a stand-alone.  I suspect that the Network Layer result type is only valid within ArcGIS Pro.  The wording in the documentation should be made more clear if this is the case.

At the moment, this isn't a useful option for me as the stand-alone script gets published as a GP service so that it can be called from a web app.