Overly fancy sample technique -- could just use `rbinom` rather than `sample` here

rbinom_with_sample.R

``` r
library(tidyverse)
tibble(x = runif(100), x_inv = 1 - x) %>%   
  mutate(event = map2_lgl(.x = x, 
                          .y = x_inv, 
                          ~sample(c(TRUE, FALSE), 
                                  size = 1, 
                                  replace = TRUE, 
                                  prob = c(.x, .y))
                          )
  )
#> # A tibble: 100 x 3
#>         x  x_inv event
#>     <dbl>  <dbl> <lgl>
#>  1 0.0366 0.963  FALSE
#>  2 0.280  0.720  FALSE
#>  3 0.475  0.525  TRUE 
#>  4 0.151  0.849  FALSE
#>  5 0.978  0.0222 TRUE 
#>  6 0.322  0.678  FALSE
#>  7 0.109  0.891  FALSE
#>  8 0.722  0.278  TRUE 
#>  9 0.273  0.727  FALSE
#> 10 0.717  0.283  FALSE
#> # ... with 90 more rows
```

<sup>Created on 2019-02-06 by the [reprex package](https://reprex.tidyverse.org) (v0.2.1)</sup>