Appending Existing Feature Service using detect_objects ArcGIS API for Python

Question

In the official documentation, detect_object said to be able to use existing Feature Service for the output_name variable.

output_name

Optional. If not provided, a FeatureLayer is created by the method and used as the output . You can pass in an existing Feature Service Item from your GIS to use that instead. Alternatively, you can pass in the name of the output Feature Service that should be created by this method to be used as the output for the tool. A RuntimeError is raised if a service by that name already exists

However, I got this error

---------------------------------------------------------------------------
AttributeError                            Traceback (most recent call last)
In  [30]:
Line 1:     out_objects = detect_objects(

File D:\Program Files\ArcGIS\Pro\bin\Python\envs\arc-big\Lib\site-packages\arcgis\learn\__init__.py, in detect_objects:
Line 347:   return gis._tools.rasteranalysis.detect_objects_using_deep_learning(

File D:\Program Files\ArcGIS\Pro\bin\Python\envs\arc-big\Lib\site-packages\arcgis\_impl\tools.py, in detect_objects_using_deep_learning:
Line 11600: output_service_name = output_objects.replace(" ", "_")

File D:\Program Files\ArcGIS\Pro\bin\Python\envs\arc-big\Lib\site-packages\arcgis\gis\__init__.py, in __getattr__:
Line 13557: raise AttributeError(

AttributeError: 'Item' object has no attribute 'replace'
---------------------------------------------------------------------------

MRE

from arcgis import GIS
from arcgis.learn import Model, detect_objects

gis = GIS(
    url='url', # fill with correct credentials
    username='username',
    password='username',
)

detect_objects_model_package = gis.content.get('some_id')
raster = gis.content.get('some_other_id')

detect_objects_model = Model(detect_objects_model_package)
detect_objects_model.install(gis=gis)

test_notebook = gis.content.get('another_id') # this is the target feature service

context = {
    'cellSize': 0.1,
    'processorType':'GPU'
}

out_objects = detect_objects(
    input_raster=raster,
    model=detect_objects_model,
    output_name=test_notebook,
    context=context,
    gis=gis
)

Questions

My goal is to append the inferencing result to an existing feature service

1. Given the error, is it correct for me to assume that currently the variable only accepts string as input?
   1. If it is, is it possible to append the inferencing result (assume that I have the item id for the existing feature service I want to append) to the existing feature service using detect_objects only? (I have the work around with .edit_features(), but to me the documentation implies it can use an existing feature service as the output, so I'm curious)
   2. If it is not (i.e. it can actually use another kind of input), what kind of input do I need to put to make it append the result to an existing feature service

Thanks!

Answer 1

Note: Not a fix -- just wanted to post this to highlight how you can further investigate the traceback by inspecting ArcGIS API for Python source to confirm/deny your suspicions.

---

Yes, looking at the detect_objects() function in version 2.3.1 of the ArcGIS API for Python, it only has logic to handle str or None input for output_name.

Inspecting arcgis/learn/__init__.py > detect_objects() reveals that it just calls gis._tools.rasteranalysis.detect_objects_using_deep_learning() and arcgis/_impl/tools.py > _RasterAnalysisTools.detect_objects_using_deep_learning() (where the error occurs) only handles None case, otherwise handling object like str:

if output_objects is None:
    output_service_name = (
        "DetectObjectsUsingDeepLearning_" + _id_generator()
    )
    output_objects = output_service_name.replace(" ", "_")
else:
    output_service_name = output_objects.replace(" ", "_")

Looks like you already submitted an issue for this: https://github.com/Esri/arcgis-python-api/issues/1947. Not much else to do for now, unless you want to edit source which you could do by updating arcgis/_impl/tools.py > _RasterAnalysisTools.detect_objects_using_deep_learning() to handle an arcgis.gis.Item object.