RXR-2949: Clean up package

[mccarthy-m-g]

This PR does some package clean up, specifically:

• updated README + using README Rmd
• using codecov in CI
• using pkgdown
• adding a package doc
• cleaning up function code and improving documentation
• adding tests
• adding a NEWS file
• adding MIT license

There are currently some outstanding TODOs, so I've marked this as a draft:

• implementing codecov and pkgdown on GitHub (I either need repo permissions or you'll need to set up the remaining bits)
• write tests for reformat_data_sdtm_to_modeling(). This function is currently a big monolith. It would be nice to refactor into smaller functions for readability and testing
• write code for reformat_data_modeling_to_nca() + tests. This function currently exists but has no code within the function
• several misc TODOs (search TODO in the project)

[roninsightrx]

implementing codecov and pkgdown on GitHub (I either need repo permissions or you'll need to set up the remaining bits)

I added the "DS team" as admin member in this repo, you should be able to do anything now

write tests for reformat_data_sdtm_to_modeling(). This function is currently a big monolith. It would be nice to refactor into smaller functions for readability and testing

I would leave this for now, it's not currently being used in production. (can mark with a TODO, and maybe make a dummy tests file or so).

write code for reformat_data_modeling_to_nca() + tests. This function currently exists but has no code within the function
several misc TODOs (search TODO in the project)

Same, not used in production, can be left for now. Another alternative could be to remove these functions from the main branch, but add them to a dev branch.

[roninsightrx]

all updates lgtm!

[roninsightrx]

@mccarthy-m-g FYI, changed name of the iterations argument in sample_covariates_mice(). replicates better matches what the output is, it's not really iterations like in fitting, it's actually creating longer output when not 1. Also in the mice package it's called number of replicates, not iterations.