Adding examples for Encoding Module

[datacubeR]

Hi @solegalli

I just added examples for the encoding module.

• I would like your feedback on this especially because I'm not sure how to test this examples are ok. Is there any way to render a dev documentation to see it?
• Second, scikit-learn shows some kind of glimpse of the answer, but they use np.arrays so it's a bit more concise than showing a glimpse of a dataframe, what do you think about that?

Please let me know if we are OK with this before start working on other modules.

Finally, do you think it could be a good a idea to have some notebooks for every module with this examples?

Thanks,

Alfonso

[solegalli]

Hi @datacubeR

great job, thank you!

Here are guidelines to build the docs. After these 2 commands, you should see a folder called "build" and in there, you can double click on the file called "index" to render the docs in a browser.

I think, the only thing I would add is a display of the transformed dataframe? I think that may help users to visually understand what the output of the encoding is meant to be.

We do have notebooks with examples in this repo: https://github.com/feature-engine/feature-engine-examples

And we link to them from the user guide. Not sure if they need updating after the various releases of Feature-engine.

Also, let's make one PR per module (this was probably your idea already?). It helps with the review and also to link with other PRs that may modify some of the same files.

Thanks a lot for your contribution!

[datacubeR]

@solegalli
I rendered a local vervion of the docs to check how it is being rendered. I get this:

So it can be seen that >>> doesn't work like in Scikit-Learn (see this) allowing to hide the prompts.

I think Scikit-Learn uses something like this.

I found this site with sphinx extensions https://sphinx-extensions.readthedocs.io/en/latest/sphinx-prompt.html, but no sure if you would like to add something like this or just get rid of the >>>.

[solegalli]

Hi @datacubeR

Yes, I am aware that you need to add some functionality to be able to remove the >>>.

Could you create an issue with the link?

I'd say, leave the >>> in, and we'll take care of adding the functionality to remove them, in a different PR.

Thank you!

[solegalli]

Or actually, when you are done with the code for this PR, I can try adding the functionality, before merging, because it seems fairly straightforward.

Let me know when you finish here, and I can give it a go.

[codecov]

Codecov Report

Merging #547 (fc40473) into main (50cecc5) will not change coverage.
The diff coverage is n/a.

@@           Coverage Diff           @@
##             main     #547   +/-   ##
=======================================
  Coverage   97.47%   97.47%           
=======================================
  Files          90       90           
  Lines        3480     3480           
  Branches      681      681           
=======================================
  Hits         3392     3392           
  Misses         32       32           
  Partials       56       56           

Impacted Files	Coverage Δ
feature_engine/encoding/count_frequency.py	96.87% <ø> (ø)
feature_engine/encoding/decision_tree.py	96.07% <ø> (ø)
feature_engine/encoding/mean_encoding.py	100.00% <ø> (ø)
feature_engine/encoding/one_hot.py	100.00% <ø> (ø)
feature_engine/encoding/ordinal.py	97.22% <ø> (ø)
feature_engine/encoding/rare_label.py	100.00% <ø> (ø)
feature_engine/encoding/similarity_encoder.py	98.63% <ø> (ø)
feature_engine/encoding/woe.py	100.00% <ø> (ø)

📣 We’re building smart automated test selection to slash your CI/CD build times. Learn more

[datacubeR]

@solegalli this one is ready for review. If OK, I can move on with other modules (in different PRs).

[solegalli]

Hi @datacubeR

Yes, this is good to merge.

I made a PR to your repo: datacubeR#1 to add the button to remove the >>>

If you merge that PR, then I go ahead and merge here.

I look forward to your examples for the other modules :)

Thanks a lot!

[datacubeR]

Hi @solegalli quick question, to check the resulting copy button. My local branch is now 1 commit behind because of this PR merge. What is the proper way to update local branch with this updated one? Is it just a rebase? how should that be applied?

[solegalli]

You mean to sync your feature branch with the PR I made to your repo?

I think, after you merge the PR, if you do git pull origin feature-branch-name should work. But you won't need that anymore, because I am merging as we speak :)

Thanks a lot for the contribution @datacubeR !