decrease default bootstrap to make examples run faster

benkeser · May 20, 2021 · 5838419 · 5838419
1 parent 3d3f434
commit 5838419
Show file tree

Hide file tree

Showing 11 changed files with 16 additions and 16 deletions.
diff --git a/R/cdf_estimation.R b/R/cdf_estimation.R
@@ -136,7 +136,7 @@ drord <- function(
   param = c("weighted_mean", "log_odds", "mann_whitney"),
   ci = "wald",
   alpha = 0.05, 
-  nboot = 1e4,
+  nboot = 1e3,
   return_models = TRUE, 
   est_dist = TRUE, 
   stratify = FALSE, 

diff --git a/README-unnamed-chunk-3-1.png b/README-unnamed-chunk-3-1.png
diff --git a/README-unnamed-chunk-4-1.png b/README-unnamed-chunk-4-1.png
diff --git a/docs/articles/using_drord.html b/docs/articles/using_drord.html
diff --git a/docs/articles/using_drord_files/figure-html/cdf-plot-1.png b/docs/articles/using_drord_files/figure-html/cdf-plot-1.png
diff --git a/docs/articles/using_drord_files/figure-html/pmf-plot-1.png b/docs/articles/using_drord_files/figure-html/pmf-plot-1.png
diff --git a/docs/articles/using_drord_files/figure-html/pmf-plot-with-text-1.png b/docs/articles/using_drord_files/figure-html/pmf-plot-with-text-1.png
diff --git a/docs/pkgdown.yml b/docs/pkgdown.yml
@@ -3,5 +3,5 @@ pkgdown: 1.6.1
 pkgdown_sha: ~
 articles:
   using_drord: using_drord.html
-last_built: 2021-05-20T20:01Z
+last_built: 2021-05-20T20:30Z
 
diff --git a/docs/reference/drord.html b/docs/reference/drord.html
diff --git a/man/drord.Rd b/man/drord.Rd
diff --git a/vignettes/using_drord.Rmd b/vignettes/using_drord.Rmd
@@ -216,7 +216,7 @@ fit3$out_mod$treat1
 
 ## Bootstrap confidence intervals
 
-By default, `drord` will used closed-form inference. While this is generally faster, we may instead prefer to use bootstrap confidence intervals, which should provide more robust behavior when working models are misspecified. Bootstrap intervals can be requested via the `ci = "bca"`, with option `nboot` determining the number of bootstrap resamples performed. The bootstrap employed by the package is the bias corrected and accelerated bootstrap [@efron1994introduction]. **In practice, we recommend using many bootstrap samples** (e.g., the package default, `nboot = 1e4`), so that inference is not dependent on the random seed that is set; however, here we illustrate using a small number of bootstrap samples for fast compilation of this vignette. To increase speed, we also turn off estimation of confidence intervals for the CDF and PMF by setting `est_dist = FALSE`. 
+By default, `drord` will used closed-form inference. While this is generally faster, we may instead prefer to use bootstrap confidence intervals, which should provide more robust behavior when working models are misspecified. Bootstrap intervals can be requested via the `ci = "bca"`, with option `nboot` determining the number of bootstrap resamples performed. The bootstrap employed by the package is the bias corrected and accelerated bootstrap [@efron1994introduction]. **In practice, we recommend using many bootstrap samples** (e.g., `nboot = 1e4`), so that inference is not dependent on the random seed that is set; however, here we illustrate using a small number of bootstrap samples for fast compilation of this vignette. To increase speed, we also turn off estimation of confidence intervals for the CDF and PMF by setting `est_dist = FALSE`. 
 
 ```{r}
 (fit4 <- drord(out = covid19$out, treat = covid19$treat,