Doc for extra modules / Download datasets

[mrfmendonca]

1. Updated the documentation:

• Added docstrings to the toy_dataset_corr.py and model_utils.py
• Added a page for both modules in the docs
• Added a class diagram to the docs of each class. This is a graphical way of showing the hierarchy between the classes

2. Updated all notebooks that uses any dataset in the datasets/ folder

• Now the datasets are downloaded in execution time. This way, the user doesn't have to download any dataset prior to executing the notebooks
• Removed the datasets/ folder from the repo, since this folder is now created dynamically when the dataset is downloaded

[codecov[bot]]

The author of this PR, mrfmendonca, is not an activated member of this organization on Codecov. Please activate this user on Codecov to display this PR comment. Coverage data is still being uploaded to Codecov.io for purposes of overall coverage calculations. Please don't hesitate to email us at support@codecov.io with any questions.

[danyrouh]

Is the _build folder under the docs folder required? Too many binary files. Is that how the docs tool works? Maybe, check in the code and the means to build it?

[markenc]

Yes, that works, thanks!

- My working hours may be different from yours. Please don't feel you need to reply outside your normal working hours.

From: Matheus Mendonça @.> Sent: Tuesday, July 19, 2022 6:46 AM To: microsoft/responsible-ai-toolbox-mitigations @.> Cc: Mark Encarnación @.>; Review requested @.> Subject: Re: [microsoft/responsible-ai-toolbox-mitigations] Doc for extra modules / Download datasets (PR #25)

@mrfmendonca commented on this pull request.

---

In README.mdhttps://nam06.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2Fmicrosoft%2Fresponsible-ai-toolbox-mitigations%2Fpull%2F25%23discussion_r924523627&data=05%7C01%7Cmarkenc%40microsoft.com%7C9c062759fb3d4d08383408da698cfe8e%7C72f988bf86f141af91ab2d7cd011db47%7C1%7C0%7C637938351549491098%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000%7C%7C%7C&sdata=S6hQNCPUE4YjmROsaoQl9R6ymeujRp9CphGOq0VQcYY%3D&reserved=0:

@@ -8,6 +8,17 @@ There are two main functions of this library:

The goal of this module is to provide a unified interface for different mitigation methods scattered around

multiple machine learning libraries, such as scikit-learn, mlxtend, sdv, among others.

+## Installation

+

+To install this package, follow these steps:

+- clone this repo

+- move to the root folder of the repo

+- install the package using:

+```console

+> pip install -e .

+```

+

+NOTE: The installation process will be updated when the package is released publicly in PyPi.

For now, I'm using the "Issues" tab to keep track of what needs to be updated. I even asked Morris to open new issues whenever he found a problem while he is testing the library. Some of the fixes in this PR are based on comments from Morris that can be found in the "Issues" section. I think we could use this approach. What do you think? Or we can also use a separate tool to keep track of the TODO list.

- Reply to this email directly, view it on GitHubhttps://nam06.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2Fmicrosoft%2Fresponsible-ai-toolbox-mitigations%2Fpull%2F25%23discussion_r924523627&data=05%7C01%7Cmarkenc%40microsoft.com%7C9c062759fb3d4d08383408da698cfe8e%7C72f988bf86f141af91ab2d7cd011db47%7C1%7C0%7C637938351549491098%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000%7C%7C%7C&sdata=S6hQNCPUE4YjmROsaoQl9R6ymeujRp9CphGOq0VQcYY%3D&reserved=0, or unsubscribehttps://nam06.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2Fnotifications%2Funsubscribe-auth%2FALUDIRPNJIF2REMZGMHWPKLVU2WQ7ANCNFSM54AAJ33A&data=05%7C01%7Cmarkenc%40microsoft.com%7C9c062759fb3d4d08383408da698cfe8e%7C72f988bf86f141af91ab2d7cd011db47%7C1%7C0%7C637938351549491098%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000%7C%7C%7C&sdata=YVA3ugp7BcB9fQqbv2P9HlphihtxFvpGEq9iLG66Ias%3D&reserved=0. You are receiving this because your review was requested.Message ID: @.**@.>>

[mrfmendonca]

Is the _build folder under the docs folder required? Too many binary files. Is that how the docs tool works? Maybe, check in the code and the means to build it?

Yeah, the amount of binary files here is bothering me as well. I really don't know the answer, since I don't know exactly what is needed to maintain the documentation webpage. I'll take a look at this.

[danyrouh]

Is the _build folder under the docs folder required? Too many binary files. Is that how the docs tool works? Maybe, check in the code and the means to build it?

Yeah, the amount of binary files here is bothering me as well. I really don't know the answer, since I don't know exactly what is needed to maintain the documentation webpage. I'll take a look at this.

• It is a tough one if that's how the docs work.
• There are also few .csv data files. I thought you download those as needed.
• I will probably clear the output of your notebooks before checking them in.

[mrfmendonca]

I updated the docs/ folder by removing the _build/ folder and adding this folder to .gitignore. In the end, we don't need the binary files, since we recompile the docs in the github-pages.yml workflow.