Add generated quantities to Stan files

[benjaminhlina]

• Adding generated quantities to stan files for each model - these allow for posterior predictive sampling to occur which gives the model the ability to validated against the raw counts of observations at each station.
• coa
• coa_time_varing
• coa_tag_integrated

I know this will likely break a bunch of stuff but I had a hard time implementing things without changing naming schemes a bit. I basically changed everything to snake_case as I often have a hard time reading other formats/get confused. If you would like this in a different PR that's fine I can change that.

I have also renamed COA_standard() function and Stan file to COA. If you'd prefer that to occur in a different PR that is cool with me as well. I know we talked about needing to keep COA_standard() ect. and depreciate them instead of making a hard change like this. Again that can be done in a different PR.

[benjaminhlina]

I also have made a utils file but that isn't in the PR - my thought on the utils file is that all the arguments that currently exist in coa will be hidden from the user and all the preprocessing e.g. making the three-demensional detection array will be a function called in coa() that is in the utils file. We could keep it seperate if you'd like and expose these functions to the user and have intermediate steps where we have the user create these objects and then feed coa() but my thought here is that if we make things so that the user only has to create a data.frame from their existing OTN or GLATOS data with standard names that will be easier for the user to pick up.

[benjaminhlina]

Scratch this PR - I'm gonna do things right instead of quick - I'm gonna finish tests on old functions create a PR - then deperecriate those functions and create the new coa() with snake_case and tests for it all on it's own.