Benchmark QA support in the SDK #216
Conversation
|
As discussed, exposing the settings that you can choose on startup would be the right thing to do (can be just an enum and then send over a set/list of the enum entries). (I'm assuming here that all our settings that we want to expose are boolean, that might not be the case) |
alexey-cord-tech
force-pushed
the
ak/shadow-data-filter-and-view
branch
from
e16c3ff
to
ea64bc0
Compare
There was a problem hiding this comment.
One comment left for somewhere to clarify. Also, there will be a project.rst which walks you through the project creation. Do you think a sentence about project modes and a link maybe to the relevant in code docstrings would be worth there?
|
I would wait with updating the "scenario" documentation for a bit, and when update it once we build the "class weights" thing and expose it here. Right now this is mostly a Tractable-specific thing and we can just point them to the right place in the docs. |
There was a problem hiding this comment.
Thanks for being so thorough here. It's fine for now if you're in a rush (however you'll need for the BE to roll out anyways).
I'd still recommend you going here to see that it would be nice to add only 2 bullet points, one with the ontology_hash, one with the workflow_settings. They can pretty much link to the new objects you have defined nicely.
You will find the file in code under docs/source/tutorials/project.rst
Introduction and Explanation
Introducing support for 'shadow data' in the SDK.
JIRA
https://cord-team.atlassian.net/browse/DEV-2055
Documentation
Not quite.
Tests
SDK tests to follow
Known issues
I'm not entirely sure about the 'ProjectWorkflowType' thing.