You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
Missing CI/CD and tests for contributed code in AutoML/census_income_prediction #2380
It looks like someone is circumventing our style checks and avoiding contributing tests by putting code in Jupyter notebooks.
See this PR for additional details regarding the missing tests for AutoML/census_income_prediction. For all functions within the code, there should be an accompanying test and the contributed Python should meet the style standards for our repo.
There are definitely errors in the code that would be caught by lint, for example, most of the code uses 2 spaces where there should be 4 for indentations. Another example, on lines 288-289, it looks like the import order is wrong. There are probably a lot more but it's difficult to tell with the code obfuscated within Jupyter markdown.
Activity
andrewferlitsch commentedon Oct 7, 2019
@sirtorry Can you take a look at this issue.
sirtorry commentedon Oct 7, 2019
Working on it this week.
andrewferlitsch commentedon Dec 2, 2019
@sirtorry Any update on the issue?
czahedi commentedon Jan 15, 2020
Hi Torry, gentle reminder ping about this issue. Thanks!
sirtorry commentedon Jan 16, 2020
This quarter, we'll be introducing a notebook creation process that will address testing. At which point, we'll resolve this issue.