Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

[ENH] Examples for YtoX transformer #6028

Merged
merged 2 commits into from Mar 10, 2024
Merged

[ENH] Examples for YtoX transformer #6028

merged 2 commits into from Mar 10, 2024

Conversation

fkiraly
Copy link
Collaborator

@fkiraly fkiraly commented Feb 28, 2024

Adds examples for YtoX transformer in its docstring; also improves docstring formatting and clarity.

@fkiraly fkiraly added documentation Documentation & tutorials module:transformations transformations module: time series transformation, feature extraction, pre-/post-processing labels Feb 28, 2024
@geetu040
Copy link
Contributor

geetu040 commented Mar 1, 2024

@fkiraly That is an easy fix, if you ask me I'll be happy to check, test, fix and add the examples for this class.

@fkiraly
Copy link
Collaborator Author

fkiraly commented Mar 1, 2024

Would you like to add more examples? Sure, if you would like to, in anothe PR?

@geetu040
Copy link
Contributor

geetu040 commented Mar 1, 2024

Thanks, I'll do in another PR

@fkiraly
Copy link
Collaborator Author

fkiraly commented Mar 1, 2024

Sure! I recommend then branching off this branch, to avoid merge conflicts.

geetu040 added a commit to geetu040/sktime that referenced this pull request Mar 4, 2024
@geetu040
Copy link
Contributor

geetu040 commented Mar 4, 2024

Thanks for letting me do the examples. I was really struggling with YtoX and thought it was irrelevant but expermenting with the examples made me rethink of its use. I previously worked on EEG dataset and just realized how easy would have the things been if I used this approach.

I have made minor fixes to examples, branching off from this branch, in this PR: #6059

@fkiraly
Copy link
Collaborator Author

fkiraly commented Mar 4, 2024

I previously worked on EEG dataset and just realized how easy would have the things been if I used this approach.

Interesting statement - what this makes me think, and imagine reading between the lines, is that the tutorials are not as clear as they could be on building pipelines from the various components.

I would be keen to hear your opinions, on where/how you specifically would have expected explanations, or whether there were places where you felt more confused than informed.

@geetu040
Copy link
Contributor

geetu040 commented Mar 4, 2024

There were actually no example use-cases for YtoX. I only found one in this discussion #5064.
I asked on discord for explanation and thankfully this PR in response made it clear for me. It was lacking dataset so it took some time for me to process, just that now I understand it.

It was simple, I was just not looking with the right approach. I just thought why to use a transformer to convert Y to X when I can just create a duplicate, process it and pass it as X. but I just realized using YtoX in a pipeline makes the code way easier and efficient.

Thanks

@fkiraly
Copy link
Collaborator Author

fkiraly commented Mar 4, 2024

There were actually no example use-cases for YtoX.

Oh - that seems like an oversight. I would have expected them in tutorial 03b.

Perhaps it's time for an update, there are various workshops (in the sktime GitHub organisation, www.github.com/sktime) which might have relevant material, in addition.

@fkiraly fkiraly merged commit 2531edb into main Mar 10, 2024
2 of 3 checks passed
@fkiraly fkiraly deleted the ytox-examples branch March 10, 2024 19:57
fkiraly pushed a commit that referenced this pull request Mar 16, 2024
#### Reference Issues/PRs

Continuation to this PR: #6028

#### What does this implement/fix? Explain your changes.

- In the mentioned PR, examples were added for `YtoX`. I have made minor
fixes in import and pipeline initialization.
- Added `load_airline` dataset in both examples.
- Replaced `ARMIAX` with `Prophet` which made the forecast resutls
better for this particular use-case.
- Completed the missing lines to make sure examples can be copied and
pasted without any error.

#### What should a reviewer concentrate their feedback on?

I have considered `pydocstyle` documentation for examples and put `#
doctest: +SKIP` for soft-dependencies. I hope I don't miss any line.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
documentation Documentation & tutorials module:transformations transformations module: time series transformation, feature extraction, pre-/post-processing
Projects
None yet
Development

Successfully merging this pull request may close these issues.

None yet

2 participants