Augment user-supplied data with fitted values and residuals from a fetwfe fit

Computes .fitted = X %*% beta_hat + x$y_mean and .resid = data[[x$response_col_name]] - .fitted, then column-binds those two columns onto data. The response mean and column name are stored on the fitted object during fitting (the estimator internally centers y before solving), so fitted values come back on the original-response scale without the caller having to remember either.

Usage

# S3 method for class 'fetwfe'
augment(x, data, ...)

Arguments

x

An object of class "fetwfe".

data

A panel data.frame with one row per unit-period (any sort order — augment auto-sorts), containing the response column under the same name used at fit time (see x$response_col_name). First-period-treated units, if present, are auto-trimmed.

...

Unused.

Value

A copy of data with two extra numeric columns: .fitted and .resid.

Details

data is auto-handled to match the fitted design: rows are auto-sorted by (unit, time), and any first-period-treated units (whose treatment effect cannot be identified by the estimator) are auto-trimmed via idCohorts(). So you can pass the same raw pdata you handed to fetwfe() — the method takes care of alignment. The only hard requirement is that data contains the response column under its original name.

Examples

if (FALSE) { # \dontrun{
  sim <- simulateData(genCoefs(G = 3, T = 6, d = 2, density = 0.5,
                               eff_size = 2),
                      N = 120, sig_eps_sq = 1, sig_eps_c_sq = 0.5, seed = 123)
  res <- fetwfeWithSimulatedData(sim)
  broom::augment(res, data = sim$pdata)
} # }