Produce local ecological estimates

Projects predictions from a fitted regression model onto the accounting constraint using a provided residual covariance matrix. This ensures that each set of local estimates satisfies the accounting identity. Local estimates may be truncated to variable bounds.

Usage

ei_est_local(
  regr,
  data,
  r_cov = NULL,
  bounds = NULL,
  conf_level = FALSE,
  unimodal = TRUE
)

# S3 method for class 'ei_est_local'
as.array(x, ...)

Arguments

regr

A fitted regression model, from ei_ridge(), or another kind of regression model wrapped with ei_wrap_model().

data

The data frame, matrix, or ei_spec object that was used to fit the regression.

r_cov

A covariance matrix of the residuals to use in projecting the local estimates onto the accounting constraint, or a list of matrices, one for each outcome variable. Defaults to the identity matrix scaled by the residual variance of regr, corresponding to orthogonal projection. Set r_cov=1 to use a degenerate covariance matrix corresponding to a (local) neighborhood model. When there are multiple outcome variables and r_cov is a matrix, it will be applied identically to each outcome.

bounds

A vector c(min, max) of bounds for the outcome, to which the local estimates will be truncated. In general, truncation will lead to violations of the accounting identity. If bounds = NULL, they will be inferred from the outcome variable: if it is contained within \([0, 1]\), for instance, then the bounds will be c(0, 1). The default bounds = FALSE forces unbounded estimates.

conf_level

A numeric specifying the level for confidence intervals. If FALSE (the default), no confidence intervals are calculated. For regr arguments from ei_wrap_model(), confidence intervals will not incorporate uncertainty in the prediction itself, just the residual. This will trigger a warning periodically.

unimodal

If TRUE, assume a unimodal residual distribution. Improves width of confidence intervals by a factor of 4/9.

x

An object of class ei_est_local

...

Additional arguments (ignored)

Value

A data frame with estimates. The .row column in the output corresponds to the observation index in the input. It has class ei_est_local, supporting several methods.

Details

Local estimates are produced independently for each outcome variable. Truncation to bounds, if used, will in general lead to estimates that do not satisfy the accounting identity.

Methods (by generic)

• as.array(ei_est_local): Format estimates an array with dimensions <rows>*<predictors>*<outcomes>. Does not work if the object has been sorted.

Examples

if (FALSE) { # \dontrun{
data(elec_1968)

spec = ei_spec(elec_1968, vap_white:vap_other, pres_dem_hum:pres_abs,
               total = pres_total, covariates = c(state, pop_urban, farm))

m = ei_ridge(spec)

ei_est_local(m, spec, conf_level = 0.95)
suppressWarnings(ei_est_local(m, spec, bounds=c(0.01, 0.2)))
} # }