Castle data

[bcallaway11]

Hi @NickCH-K,

Thanks for putting this together---it is a great resource.

I was poking at the castle data and realized that there are some variables that are available in the Mixtape version that aren't available here. The main two ones that I see are the name of the state and the region that the state is in. There are also some "junk" variables in there like state specific trend variables.

Brant

[NickCH-K]

Thanks!

If I remember right, I believe I dropped variables for a few Mixtape datasets that did not appear to be used because the package was running up against size-maximum issues for some of the package repositories (this is also why judge_fe isn't in there at all). Do the missing variables make some of the Mixtape code examples not run properly?

[bcallaway11]

I see, that makes sense.

On my end, I was trying to use this as example data in another package, and it would be useful to have the state names and region while it is less useful to have all the columns related to pre-constructed state linear trends (and there are tons of variables in it like that). But I think it's possible that dropping those extra variables might make it where some of the Mixtape examples do not run immediately. I am not sure.

Perhaps just adding back the state name would be helpful because you could construct the region easily after that.