propensity score calculation in in bootstrap workflow

[rpruim]

In https://causal-inference-in-r.netlify.app/chapters/11-outcome-model#the-bootstrap, there is the following example:

fit_ipw <- function(.split, ...) {
  # get bootstrapped data frame
  .df <- as.data.frame(.split)

  # fit propensity score model
  propensity_model <- glm(
    park_extra_magic_morning ~ park_ticket_season + park_close + park_temperature_high,
    data = seven_dwarfs_9,           ### should this be .df?
    family = binomial()
  )

  # calculate inverse probability weights
  .df <- propensity_model |>
    augment(type.predict = "response", data = .df) |>        ### newdata = .df (or omit if using .df above)
    mutate(wts = wt_att(
      .fitted,
      park_extra_magic_morning,
      exposure_type = "binary"
    ))

  # fit correctly bootstrapped ipw model
  lm(
    wait_minutes_posted_avg ~ park_extra_magic_morning,
    data = .df,
    weights = wts
  ) |>
    tidy()
}

Should the propensity scores be recalculated from each bootstrap sample (currently: data = seven_dwarfs_9)? [If it were done as here, it would be better to compute it once rather than once per resample.]

[rpruim]

Also

augment(type.predict = "response", data = .df)

Should be newdata = .df.