Ben

Here's an edge case in which `survival::strata()` behaves differently than we might like. (Excuse for rolling our own instead?) ```r a

Would we have to export it (the optmatch version of `strata()`) in order to serve our purposes, or could we get away with having it only within the optmatch namespace?...

An optmatch-specific `strata()` function could help with the program sketched in [this comment](https://github.com/markmfredrickson/optmatch/issues/161#issuecomment-845466476) to #161, which would in turn set us up to better enable "tidy data" workflows.

Regarding names, I'm not quite ready to abandon `strata()`, but `blocking()` would have the advantages of not conflicting with that existing function plus aligning with `forcing()` as proposed over in...

`within()` and `by()` clash with base functions, but `exactly()` doesn't. It's cute, too. I continue to like `block()` and/or `blocking()` as well, but it needn't be either-or... >

Nice! Left a couple minor comments on fa1b7c4 inline.

Would ditching the survival dependency as in the `remove_survival_strata` branch address the scoping problem that was noted in [this comment to #105](https://github.com/markmfredrickson/optmatch/issues/105#issuecomment-427171212) (and cited in the problem statement at the...

I think this issue can be closed. If you agree @josherrickson, please close it. (Above I had cross-referenced [a comment to \#161](https://github.com/markmfredrickson/optmatch/issues/161#issuecomment-845466476). The connection is that that comment was proposing...

1. Do we have a sense of the additional time cost of a `det()` call over and above the `solve()`? If it gets us up to the time cost of...

Update: I'm on board with either the `det()`-based branching logic Josh proposed here or categorically using `svd()` for a generalized inverse. Would like to see inversion/gen inversion of correlation rather...