[remlapmot]

• Paul found that he needed unstabilized treatment weights in his analysis for a comparison - so I guess we may as well add it?

  Adds weight.stabilized to SEQopts() (default TRUE, unchanged behaviour). With weight.stabilized = FALSE no numerator models are fit, so the treatment weights are cumulative products of 1 / denominator; LTFU and visit weights keep their numerators.

• Also corrects the selection.random / selection.prob documentation (which I can send as a separate PR if you don't want the unstab weights added).

• And improves memory footprint of bootstrapping - was collecting n which wasn't used.

• small amends to justfile

[remlapmot]

The check failures are to do with some internet access/CRAN/ Posit PPM issue, R CMD check passes fine locally:

   Status: OK

── R CMD check results ──────────────────────────────────────────────────── SEQTaRget 1.4.5.9000 ────
Duration: 3m 26.6s

0 errors ✔ | 0 warnings ✔ | 0 notes ✔

[ryan-odea]

Looks good Tom! Thanks for catching the bootstrap memory footprint.