Function reference • svrep

Convert a Survey Design Object to a Data Frame

as_data_frame_with_weights(): Convert a survey design object to a data frame with weights stored as columns

Nonresponse Adjustments

redistribute_weights(): Redistribute weight from one group to another

Creating Bootstrap Replicate Weights

as_bootstrap_design(): Convert a survey design object to a bootstrap replicate design

make_rwyb_bootstrap_weights(): Create bootstrap replicate weights for a general survey design, using the Rao-Wu-Yue-Beaumont bootstrap method

estimate_boot_reps_for_target_cv(): Estimate the number of bootstrap replicates needed to reduce the bootstrap simulation error to a target level

estimate_boot_sim_cv(): Estimate the bootstrap simulation error

Generalized Replication Method (including the Generalized Bootstrap)

Fay’s Generalized Replication Method

as_fays_gen_rep_design(): Convert a survey design object to a replication design using Fay's generalized replication method

make_fays_gen_rep_factors(): Form replication factors using Fay's generalized replication method

Generalized Bootstrap

as_gen_boot_design(): Convert a survey design object to a generalized bootstrap replicate design

make_gen_boot_factors(): Creates replicate factors for the generalized survey bootstrap

Variance Estimators Available for the Generalized Replication Methods

The generalized replication methods (Fay’s method and the generalized bootstrap) work by “mimicking” a target variance estimator, such as the Horvitz-Thompson estimator. This help page describes the variance estimators that can be used as the target for the generalized replication methods.

variance-estimators: Variance Estimators

Helper Functions for Working with Quadratic Forms

These functions help the user specify the quadratic form representation of common variance estimators and, if necessary, adjust them so that they are positive semidefinite (a necessary prerequisite for using the generalized replication methods).

get_design_quad_form(): Determine the quadratic form matrix of a variance estimator for a survey design object

make_quad_form_matrix(): Represent a variance estimator as a quadratic form

make_twophase_quad_form(): Combine quadratic forms from each phase of a two phase design

get_nearest_psd_matrix(): Approximates a symmetric, real matrix by the nearest positive semidefinite matrix.

is_psd_matrix(): Check whether a matrix is positive semidefinite

Creating Jackknife Replicate Weights

as_random_group_jackknife_design(): Convert a survey design object to a random-groups jackknife design

Calibrating to Estimated Control Totals

calibrate_to_estimate(): Calibrate weights from a primary survey to estimated totals from a control survey, with replicate-weight adjustments that account for variance of the control totals

calibrate_to_sample(): Calibrate weights from a primary survey to estimated totals from a control survey, with replicate-weight adjustments that account for variance of the control totals

General-Purpose Helper Functions

shuffle_replicates(): Shuffle the order of replicates in a survey design object

stack_replicate_designs(): Stack replicate designs, combining data and weights into a single object

subsample_replicates(): Retain only a random subset of the replicates in a design

rescale_reps(): Rescale replicate factors

add_inactive_replicates(): Add inactive replicates to a survey design object

svyby_repwts(): Compare survey statistics calculated separately from different sets of replicate weights

Diagnostic Functions to Check Replicate Weights

summarize_rep_weights(): Summarize the replicate weights

Example Datasets

lou_pums_microdata: ACS PUMS Data for Louisville

lou_vax_survey: Louisville Vaccination Survey

lou_vax_survey_control_totals: Control totals for the Louisville Vaccination Survey

library_census library_multistage_sample library_stsys_sample: Public Libraries Survey (PLS): A Census of U.S. Public Libraries in FY2020