Random seed split

[joaorobson]

Proposed changes:

• Enable the user to set a seed to split the NLU data into test and training.
• Close Add set.seed ability to CLI when splitting NLU data #4647

Expected behavior:

• When a user is splitting the NLU data, a new argument specifying a random seed (in sklearn, known as random_state) could be passed, as shown below:
  rasa data split nlu --random-seed=42

Status (please check what you already did):

• made PR ready for code review
• added some tests for the functionality
• updated the documentation
• updated the changelog
• reformat files using black (please check Readme for instructions)

[erohmensing]

@joaorobson is this still a draft PR or is it ready for review?

[joaorobson]

@joaorobson is this still a draft PR or is it ready for review?

It is ready now. I just needed to update the docs and the changelog.

[erohmensing]

Ok, we'll give it a review as soon as possible.

[tabergma]

Looks great 🚀 Just a few comments. Thanks for adding a test!

Please, update your branch and update the changelog entry (https://github.com/RasaHQ/rasa/tree/master/changelog).

[joaorobson]

Looks great Just a few comments. Thanks for adding a test!

Please, update your branch and update the changelog entry (https://github.com/RasaHQ/rasa/tree/master/changelog).

Thanks. I guess that I've fixed everything that you requested.

[tabergma]

Thanks for the contribution! 🎉