Regression sensitivity analysis

Top-level dispatcher that mirrors the Stata regsensitivity command. For most users, calling regsen_bounds() or regsen_breakdown() directly is clearer.

Usage

regsensitivity(
  formula,
  data,
  subcommand = c("bounds", "breakdown", "summary"),
  ...
)

Arguments

formula

Two-sided formula: y ~ x + w1 + w2 + .... The first right-hand-side variable is the primary independent variable; the rest are controls.

data

A data.frame.

subcommand

One of "bounds" (the default), "breakdown", "summary".

...

Additional arguments forwarded to the underlying function.

Value

An object of class regsensitivity, or for subcommand = "summary" one of class regsensitivity_summary.

Details

Versions before 0.3.0 took subcommand first, following the Stata syntax. A call in that order, regsensitivity("bounds", y ~ x, data), still runs, with a deprecation warning.

See also

regsen_bounds(), regsen_breakdown(), regsen_summary()

Examples

data(bfg2020)
regsensitivity(
  avgrep2000to2016 ~ tye_tfe890_500kNI_100_l6 +
    log_area_2010 + lat + lon,
  data = bfg2020,
  subcommand = "bounds",
  cbar = 0.1
)
#> 
#> Regression Sensitivity Analysis -- Bounds [DMP (2026)]
#> ------------------------------------------------------------------------
#> Model:             avgrep2000to2016 ~ tye_tfe890_500kNI_100_l6
#> N (obs):           2036
#> Beta (medium):     1.3854
#> Hypothesis:        Beta > 0
#> Breakdown point:   0.3906
#> 
#> Results:
#>    rxbar rybar   cbar    bmin   bmax
#>        0  +Inf    0.1  1.3854 1.3854
#>  0.23774  +Inf    0.1 0.55995 2.2109
#>  0.47547  +Inf    0.1 -0.3236 3.0944
#>  0.71321  +Inf    0.1 -1.2923 4.0631
#>  0.95094  +Inf    0.1 -2.3864 5.1572
#>   1.1887  +Inf    0.1 -3.6709 6.4417
#>   1.4264  +Inf    0.1 -5.2616 8.0325
#>   1.6642  +Inf    0.1 -7.3957 10.167
#>   1.9019  +Inf    0.1 -10.676 13.447
#>   2.1396  +Inf    0.1  -17.46 20.231
#>   2.3774  +Inf    0.1    -Inf   +Inf
#> 
#> Use summary() for the specification and full results.