Print, summarize and plot a bootstrapped breakdown point — regsen

print() gives the point estimate and the interval regsen_boot() reported. summary() adds the analysis bootstrapped, both interval types, the BCa constants, the distribution of the replicates (their standard deviation is the bootstrap standard error) and the counts of failed and infinite replicates. plot() draws the replicates with the point estimate and the interval marked. confint() returns the interval as a matrix, recomputing it from the stored replicates when level or type differ from those of the run.

Usage

# S3 method for class 'regsensitivity_boot'
print(x, ...)

# S3 method for class 'regsensitivity_boot'
summary(object, ...)

# S3 method for class 'summary.regsensitivity_boot'
print(x, ...)

# S3 method for class 'regsensitivity_boot'
confint(object, parm, level = object$level, type = object$type, ...)

# S3 method for class 'regsensitivity_boot'
plot(x, bins = 30, base_size = 11, ...)

Arguments

x, object: A regsensitivity_boot object, or for the print() method of its summary, a summary.regsensitivity_boot object.
...: Ignored.
parm: Ignored; there is one parameter.
level: Confidence level. Defaults to the level of the run.
type: "bca" or "perc". Defaults to the interval the run reported. "bca" needs the jackknife, which a run with type = "perc" skipped.
bins: Number of histogram bins.
base_size: Base font size, passed to theme_regsen().

Value

confint() returns a 1 x 2 matrix. summary() returns an object of class summary.regsensitivity_boot. plot() returns a ggplot object.

Details

Every value is the magnitude of the breakdown point, as printed by regsen_breakdown().

Examples

# \donttest{
data(bfg2020)
f <- avgrep2000to2016 ~ tye_tfe890_500kNI_100_l6 + log_area_2010 + lat + lon
bb <- regsen_boot(f, bfg2020, compare = c("log_area_2010", "lat", "lon"),
                  R = 99, type = "perc", seed = 1)
bb
#> Bootstrap confidence interval for the breakdown point
#> ------------------------------------------------------------
#>   Point estimate     : 0.3749
#>   95% CI (percentile): [0.2774, 0.5079]
#>   Replicates         : 99
#> 
#> Use summary() for the replicate distribution and both intervals.
summary(bb)
#> Bootstrap inference for the breakdown point
#> ------------------------------------------------------------
#>   Analysis           : DMP (2026)
#>   Treatment          : tye_tfe890_500kNI_100_l6
#>   Hypothesis         : Beta > 0
#>   Replicates (R)     : 99
#>   Cluster bootstrap  : no
#> 
#> --- Estimate --------------------------------------------
#>   Point estimate     : 0.3749
#>   Bootstrap SE       : 0.0662
#>   Bootstrap bias     : 0.0032
#>   95% CI (percentile): [0.2774, 0.5079]
#> 
#> --- Finite replicates -----------------------------------
#>    Min     Q1 Median     Q3    Max 
#> 0.2086 0.3325 0.3675 0.4167 0.5812 
confint(bb, level = 0.9)
#>                 5 %      95 %
#> breakdown 0.2880669 0.4892533
plot(bb)

# }