Skip to contents

Tidy a FETWFE_tes simulation truth object

Source: R/broom_methods.R
tidy.FETWFE_tes.Rd

Returns a broom-style tidy data frame for the population-truth object returned by getTes(). Row 1 is the overall true ATT (term = "ATT_true"); subsequent rows are the true cohort ATTs (term = "Cohort <adoption-time>", using the simulator's convention that cohort g adopts at calendar time g + 1, so the labels match what tidy.<estimator> uses on a fitted panel generated from the same FETWFE_coefs). Standard error / statistic / p-value columns are always NA_real_ — there is no sampling distribution for a population truth. When conf.int = TRUE (default, matching the sibling tidy methods), conf.low / conf.high columns are included and also set to NA_real_. When conf.int = FALSE, those columns are omitted.

Usage

# S3 method for class 'FETWFE_tes'
tidy(x, conf.int = TRUE, conf.level = 0.95, ...)

Arguments

x

An object of class "FETWFE_tes" returned by getTes().

conf.int

Logical; include conf.low / conf.high columns. Defaults to TRUE to match the sibling tidy methods and preserve pre-#84 backward compatibility. Population-truth objects have no sampling distribution, so the CI columns are always filled with NA_real_ when included.

conf.level

Numeric in (0, 1). Accepted for broom-convention parity but unused (no CIs to compute for a population truth); validated regardless. Defaults to 0.95 (FETWFE_tes objects do not carry an alpha slot, so there is no fitted-object value to default to).

...

Unused.

Value

A data frame with G + 1 rows and columns term, estimate, std.error, statistic, p.value, and (when conf.int = TRUE) conf.low / conf.high.

Examples

if (FALSE) { # \dontrun{
  coefs <- genCoefs(G = 3, T = 6, d = 2, density = 0.5, eff_size = 2)
  broom::tidy(getTes(coefs))
} # }