Update documentation to include all data checks and usage of data checks in AutoML#1412
Update documentation to include all data checks and usage of data checks in AutoML#1412angela97lin merged 23 commits intomainfrom
Conversation
Codecov Report
@@ Coverage Diff @@
## main #1412 +/- ##
=======================================
Coverage 100.0% 100.0%
=======================================
Files 216 216
Lines 14228 14228
=======================================
Hits 14221 14221
Misses 7 7
Continue to review full report at Codecov.
|
angela97lin
requested review from
bchen1116,
dsherry,
eccabay,
freddyaboulton and
jeremyliweishih
bchen1116
left a comment
bchen1116
left a comment
There was a problem hiding this comment.
Docs look great! Left a comment on OutlierDataCheck but everything else LGTM
dsherry
left a comment
dsherry
left a comment
There was a problem hiding this comment.
LGTM! Left a few suggested tweaks
| "\n", | ||
| "* if any of the target values are missing, an error is returned\n", | ||
| "* if the specified problem type is a binary classification problem but there is more or less than two unique values in the target, an error is returned\n", | ||
| "* if binary classification target classes are numeric values not equal to {0, 1}, an error is returned because it can cause unpredictable behavior when passed to pipelines" |
There was a problem hiding this comment.
This is great
Is the last point still true? Now that we have the label encoder in the classification pipeline, I think this use case will work fine. If I'm right, can we file something to either delete this from InvalidTargetDataCheck, or change it to a warning instead of an error? Certainly, providing something which looks suspicious could indicate the user has provided incorrect data for some reason, which makes me think showing a warning is a good idea rather than just deleting.
There was a problem hiding this comment.
Hmmm, good point! I just filed #1422 to track investigating and updating the docs accordingly. For now, I'll merge as is with the documentation since that's the current behavior.
freddyaboulton
left a comment
freddyaboulton
left a comment
There was a problem hiding this comment.
@angela97lin Nice! I left some minor comments.
4 participants
Closes #1395
Updated docs: Data checks
Data checks in AutoML