Creates a svyder object from pre-computed DER pipeline results.
This is the internal constructor; users should call der_compute()
or der_diagnose() instead.
Usage
new_svyder(
der,
params,
H_obs,
J_cluster,
V_sand,
sigma_mcmc,
deff_j,
B_j,
classification,
tau,
corrected_draws,
scale_factors,
original_draws,
call,
family,
n_obs,
n_groups,
compute_time
)Arguments
- der
Named numeric vector of Design Effect Ratios.
- params
Character vector of parameter names.
- H_obs
Numeric matrix, observed information matrix (d x d).
- J_cluster
Numeric matrix, clustered score outer product (d x d).
- V_sand
Numeric matrix, sandwich variance (d x d).
- sigma_mcmc
Numeric matrix, MCMC posterior covariance (d x d).
- deff_j
Numeric vector of cluster-level design effects (length J).
- B_j
Numeric vector of shrinkage factors (length J).
- classification
Data frame with per-parameter classification results.
- tau
Numeric scalar, flagging threshold for DER.
- corrected_draws
Numeric matrix, corrected posterior draws (M x d).
- scale_factors
Numeric vector of Cholesky scale factors (length d).
- original_draws
Numeric matrix, original posterior draws (M x d).
- call
The matched call that produced this object.
- family
Character string, model family (e.g., "binomial", "gaussian").
- n_obs
Integer, number of observations.
- n_groups
Integer, number of groups/clusters.
- compute_time
Numeric scalar, computation time in seconds.