Fix notebooks in inference track #974
Conversation
|
Check out this pull request on See visual diffs & provide feedback on Jupyter Notebooks. Powered by ReviewNB |
|
🚀 Deployed on https://deploy-preview-974--etna-docs.netlify.app |
|
View / edit / reply to this conversation on ReviewNB martins0n commented on 2022-10-07T06:04:27Z Line #6. from etna.models.mixins import MultiSegmentModelMixin Let's dont use these mixins, these are technical once for own use. I guess we should make example from scratch without magic and multiple dependencies Mr-Geekman commented on 2022-10-07T07:07:23Z Even for per-segment type? |
|
Even for per-segment type? View entire conversation on ReviewNB |
This reverts commit 3c2f845.
Codecov Report
@@ Coverage Diff @@
## inference #974 +/- ##
============================================
Coverage ? 85.24%
============================================
Files ? 132
Lines ? 7708
Branches ? 0
============================================
Hits ? 6571
Misses ? 1137
Partials ? 0 📣 We’re building smart automated test selection to slash your CI/CD build times. Learn more |
|
View / edit / reply to this conversation on ReviewNB martins0n commented on 2022-10-07T11:58:29Z These classes have different signatures for
methods theirs
if a model
It is an interval of dataset that goes right before the interval we want to make prediction on.
It is a part of a dataset before predection points that is necessary for making forecasts
predictions into the future - > seems like it's about any predicitions |
|
View / edit / reply to this conversation on ReviewNB martins0n commented on 2022-10-07T11:58:30Z Line #18. ) What about kwargs and parameters saving for to_dict |
|
View / edit / reply to this conversation on ReviewNB martins0n commented on 2022-10-07T11:58:30Z Line #2. Why do we need this if we use pipeline below anyway? |
|
View / edit / reply to this conversation on ReviewNB martins0n commented on 2022-10-07T11:58:31Z predictions have been made - maybe |
|
View / edit / reply to this conversation on ReviewNB martins0n commented on 2022-10-07T11:58:32Z Maybe we have the difference becouse of randomization? Mr-Geekman commented on 2022-10-07T12:08:48Z I will fix random seed inside model. |
.binder/requirements.txt
Outdated
| @@ -1 +1 @@ | |||
| etna[all] @ git+https://github.com/tinkoff-ai/etna.git@master | |||
| etna[all] @ git+https://github.com/tinkoff-ai/etna.git@issue-974 | |||
There was a problem hiding this comment.
I will remove it, it is for the case when I want to run it on clean machine in binder (some warnings aren't present where).
|
I will fix random seed inside model. View entire conversation on ReviewNB |
|
View / edit / reply to this conversation on ReviewNB martins0n commented on 2022-10-10T11:27:30Z Line #7. log = LogTransform(in_column="target", out_column="log") We don't use it in transforms |
This reverts commit 1dd91f6.
Before submitting (must do checklist)
Proposed Changes
Look #973.
Closing issues
Closes #973.