surveycore (version 0.8.3)

clean: Tidy a Survey GLM Fit

Description

Converts a survey_glm_fit object into a survey_glm_tidy result tibble with one row per model coefficient (plus optional reference rows for factor predictors), design-based standard errors, confidence intervals, and structured metadata.

Usage

clean(
  model,
  conf_level = 0.95,
  include_reference = TRUE,
  n = FALSE,
  statistic = TRUE,
  exponentiate = FALSE,
  interaction_sep = " * ",
  ...
)

Value

A survey_glm_tidy object: a tibble with S3 class c("survey_glm_tidy", "survey_result", "tbl_df", "tbl", "data.frame"). Metadata is accessed via meta().

Arguments

model

A survey_glm_fit object from survey_glm().

conf_level

Numeric scalar in (0, 1). Confidence level for confidence intervals. Default 0.95.

include_reference

Logical. If TRUE, reference levels for unordered factor predictors appear as rows with estimate = NA and reference_row = TRUE. Default TRUE.

n

Logical. If TRUE, adds an n_obs column with the unweighted observation count per term. Default FALSE.

statistic

Logical. If TRUE (default), includes the statistic (t-statistic) column. Set to FALSE to drop it.

exponentiate

Logical. If TRUE, exponentiates estimate, conf_low, and conf_high. std_error is left on the log scale (matching broom convention). Fires surveycore_warning_exponentiate_nonlog when the model link is not log-based. Default FALSE.

interaction_sep

Character scalar. Separator for interaction term labels. Default " * ".

...

Currently unused.

See Also

survey_glm() to fit the model, meta() to access metadata.

Other analysis: get_anova(), get_corr(), get_covariance(), get_diffs(), get_freqs(), get_means(), get_pairwise(), get_quantiles(), get_ratios(), get_t_test(), get_totals(), get_variance(), meta()

Examples

d <- as_survey(gss_2024, ids = vpsu, weights = wtssps,
               strata = vstrat, nest = TRUE)
fit <- survey_glm(d, age ~ sex)
clean(fit)
clean(fit, conf_level = 0.99, exponentiate = FALSE)