Skip to contents

Declarative estimation over a coordinated panel object

Source: R/estimation-spec.R
step_domain.Rd

step_domain() and step_estimate() build an estimation pipeline (weightflow_estimation) on top of a saved wave_bootstrap() / wave_jackknife() object, so that changes, levels and linear contrasts – with the honest overlap covariance – can be requested declaratively and disaggregated, the way the weighting recipe is built with step_*. It is a thin front end over change_estimate(), panel_estimate() and level_estimate(); the coordinated replicates do the variance. Evaluate with collect_estimates() (printing does it too).

Usage

step_domain(x, ...)

step_filter(x, condition)

step_estimate(
  x,
  statistic,
  over = NULL,
  type = c("absolute", "relative"),
  waves = NULL,
  contrast = NULL,
  level = 0.95,
  label = NULL
)

step_transition(x, from, to, format = c("row", "col", "joint", "counts"))

Arguments

x

a weightflow_wave_boot / weightflow_wave_jack, or a weightflow_estimation to extend.

...

for step_domain(), one or more grouping columns (unquoted or as strings); they stack, so the disaggregation is their cross (e.g. region x sex). A column may not be named after one of the result table's own columns (estimand, over, type, estimate, se, ci_lower, ci_upper, rho); rename it in the wave data first.

condition

for step_filter(), a logical expression (unquoted) that selects the subpopulation to estimate over – e.g. age >= 25 & age <= 54. It is evaluated in each wave's data; rows are masked (not dropped), so the coordinated replicate structure and the overlap covariance are preserved. Multiple step_filter() calls stack (their conditions are ANDed).

statistic

the estimand, as a DSL call – mean(var), total(var), prop(cond), ratio(num, den), quantile(var, p) – or a function(w, data). Every argument is evaluated in the wave data and must give one numeric or logical value per unit: prop() takes a condition (prop(status == "unemployed")), not a category, and p in quantile() is a probability in [0, 1] (0.5, not 50). ratio() is taken over the domain where numerator and denominator are both observed, as survey::svyratio(na.rm = TRUE) is. The weighted quantile is R's type 5.

over

what to estimate: "change" (between two waves), "level" (one wave) or "contrast" (a linear combination, with contrast=). Default: "change" if the object has >= 2 waves, else "level"; "contrast" is implied when contrast= is given.

type

for over = "change", "absolute" (default) or "relative" (theta2/theta1 - 1).

waves

optional wave label(s): one for "level", two for "change".

contrast

numeric weights (one per wave) for over = "contrast".

level

confidence level.

label

optional name for the estimand (defaults to the statistic's expression).

from, to

for step_transition(), the from-/to-state columns.

format

transition table format ("row", "col", "joint", "counts").

Value

a weightflow_estimation.

See also

collect_estimates(), change_estimate(), panel_estimate(), level_estimate()

Examples

t1 <- subset(panel_ine, wave == 1 & disposition == "R")
t2 <- subset(panel_ine, wave == 2 & disposition == "R")
wb <- wave_bootstrap(
  list(T1 = weighting_spec(t1, base_weights = pw),
       T2 = weighting_spec(t2, base_weights = pw)),
  replicates = 100, strata = "stratum", psu = "psu", seed = 1, progress = FALSE)

# net change of the unemployment rate, by region
collect_estimates(wb |> step_domain(region) |>
  step_estimate(mean(unemployed), over = "change"))
#> <weightflow estimates [bootstrap]  T1, T2>
#>          estimand   over     type     region estimate      se ci_lower ci_upper
#>  mean(unemployed) change absolute   Interior -0.04337 0.01267 -0.06821 -0.01853
#>  mean(unemployed) change absolute Montevideo  0.01580 0.01713 -0.01777  0.04937
#>     rho
#>  0.6845
#>  0.5741

# subpopulation: same change among the working-age population (rows masked, not dropped)
collect_estimates(wb |> step_filter(age >= 25 & age <= 54) |>
  step_estimate(mean(unemployed), over = "change"))
#> <weightflow estimates [bootstrap]  T1, T2>
#>          estimand   over     type estimate      se ci_lower ci_upper    rho
#>  mean(unemployed) change absolute -0.01542 0.01937 -0.05339  0.02254 0.4246