feat: ProphetModelDataset

[galenseilis]

Description

This PR was created to add Meta's (formerly Facebook's) Prophet model. It is a probabilistic modelling for time series forecasting.

Development notes

The Prophet module provides functions for serializing and deserializing models to/from JSON. Beyond that the behaviour is intended to be the same as the JSONDataset. I have subclassed from JSONDataset and overrode the _load and _save methods to use Prophet's serialization during these methods.

I have tried to emulate the tests for JSONDataset, but with a code duplicate for a trained Prophet model instead of a test file.

Checklist

• Opened this PR as a 'Draft Pull Request' if it is work-in-progress
• Updated the documentation to reflect the code changes
• Added a description of this change in the relevant RELEASE.md file
• Added tests to cover my changes

[merelcht]

Hi @galenseilis, thanks for the PR. I think this would be a great candidate to go in as an experimental dataset . This relaxes the requirements for tests. How does that sound?

[galenseilis]

Hi @galenseilis, thanks for the PR. I think this would be a great candidate to go in as an experimental dataset . This relaxes the requirements for tests. How does that sound?

Sounds good to me!

[merelcht]

Great! @galenseilis are you okay moving it to the experimental directory yourself or do you want me to jump in and make the PR ready for review?

[galenseilis]

Great! @galenseilis are you okay moving it to the experimental directory yourself or do you want me to jump in and make the PR ready for review?

Hi @merelcht , I would appreciate it if you would jump in to make the PR ready for review. My apologies for any inconvenience.

[merelcht]

@galenseilis no worries, happy to help!

[merelcht]

@galenseilis, I've made all changes required. I've updated the tests and docstrings. The rest are mostly docs build and linting changes. Let me know if all looks good and I'll ask the team to review 🙂

[ElenaKhaustova]

Thank you @galenseilis for the contribution and @merelcht for polishing it!

I left one nit comment but the implementation LGTM.

I approve the PR but as far as I understand it is blocked by #850 now, so we need to resolve it first.

[ankatiyar]

Thanks @galenseilis, I took the liberty to add this to the release notes and fixed bandit related security stuff! Happy to get this in for the next release!