Testing: Endpoint FSKX Model Execution API

[schuelet]

Codebase: https://github.com/KIDA-BfR/PoC-for-FSKX-API-server/blob/feature/fskx_create/fskxserver/fskxserver/fskxserver.py

Write a script that tests if all the registered models can be successfully executed. Models from RAKIP Model Repository (not all are from there) also need to run using user-defined simulation parameters.

Script should follow (and improve on) this KNIME workflow (need to be unzipped before importing to KNIME): model_execution_tests.zip

• get all models registered
  • /get-models
  • get the modelId from results
• execute all models with default parameters
  • /simulate?plot_type=plot-type can be png, svg, pdf or be left&modelId=modelId
• if modelId in https://knime.bfr.berlin/landingpage//DB/metadata/?repository=RAKIP-Web&status=Curated
  • extract parameters and values
  • /simulate?plot_type=png&modelId=modelId with those parameters
• /simulate returns task-id (simulationId)
• check status for task-id
  • /status?taskId=task-id
  • status can be RUNNING or SUCCESS (or FAILURE - check if that can happen)
• on SUCCESS:
  • get simulation result
  • /results?simulationId=task-id&filetype=plot-type Check what happens when you execute many models at the same time.

Be aware: This model will take 20 minutes, so maybe leave it out of the automatic process: { "modelId": "543d0022-a5eb-4f05-bb82-edef34ec7807", "modelName": "Listeriosis gQMRA for French population" }

[schuelet]

Example code: ` import requests

url = "https://apiserver.risk-ai-cloud.com/get-models"

payload = {} headers = { 'Authorization': '••••••' }

response = requests.request("GET", url, headers=headers, data=payload)

print(response.text) `