RxInferClient.py Example Fails with ConnectionRefusedError When Server Isn't Running

[ActiveInferenceInstitute]

Running the example Python client script support/server/RxInferClient.py fails with a ConnectionRefusedError if the corresponding RxInfer server is not running or is not accessible at the default address (http://localhost:8000/v1).

Here is the exact code I am running (side comment, having the pip install rxinferclient would be helpful in the readme).

The main issue here is that even with the localhost commented out, I am still getting connection error with the RxInferClient() issue.

# to install
# python3 -m pip install rxinferclient 

from rxinferclient import RxInferClient

# Initialize with default settings (auto-generates API key)
client = RxInferClient()

# Or initialize with custom server URL
# client = RxInferClient(server_url="http://localhost:8000/v1")

# Or initialize with your own API key
# client = RxInferClient(api_key="your-api-key")

# Ping the server to check if it's running
response = client.server.ping_server()
print(response.status)  # 'ok'

# Create a model instance
response = client.models.create_model_instance({ 
    "model_name": "BetaBernoulli-v1"
})
instance_id = response.instance_id

# Delete the model instance when done
client.models.delete_model_instance(instance_id=instance_id)

Ideally, the script should either:

• Have the base RxInferClient() method more effectively find default remote sever.
• Fail with a more specific error message clearly stating that the connection to the RxInfer server failed and indicating the required prerequisite or local steps to take.
• Or, the example script/documentation should explicitly state that a running RxInfer server instance at the specified (or default) address is required before running the client script.