fit_extract_multinom: Fit multinomial model and extract summary info on a single...
In jenniferthompson/ClusterMultinom: Clustered Multinomial Logistic Regression

Description Usage Value See Also Examples

This is a wrapper for fit_multinom() + extract_multinom_info; it fits a model and extracts summary information (indicator for successful fit, vector of coefficients, and character vector of errors/warnings) given a formula, a single data object (data.frame or mice::mids object), and a reference level for the outcome.

1	fit_extract_multinom(formula, df, ref_level, ...)

List containing the following elements:

fitsucc: logical indicating whether the model was successfully fit, with no warnings/errors
coefs: numeric vector of model coefficients (NULL if fitsucc = FALSE)
msgs: character vector of warnings/error messages (NULL if fitsucc = TRUE)
modobj: full model object of class VGAM::vglm(). Only returned if requested (coef_only = FALSE).

If df is a mice::mids object, the following elements are also returned:

nfailsucc: numeric vector; number of successful and failed model fits among imputations
impcoefs (if coef_matrix = TRUE): MxP matrix of coefficients from each individual imputation

fit_multinom, extract_multinom_info. vglm; multinomial for model fitting. mice for imputation.

my_df <- data.frame(
  id = sample(1:50, size = 500, replace = TRUE),
  x1 = sample(c(NA, 1:100), size = 500, replace = TRUE),
  x2 = sample(c(NA, 1:100), size = 500, replace = TRUE),
  y = sample(c(NA, LETTERS[1:3]), size = 500, replace = TRUE)
)

## Basic usage
my_mod_info <- fit_extract_multinom(
  formula = y ~ x1 + x2,
  df = my_df,
  ref_level = "A"
)

## Handle missingness with multiple imputation using mice
## Also try a more complicated formula that might fail to converge
my_df_mice <- mice(my_df[1:125,])
my_mod_info_mice <- fit_extract_multinom(
  formula = y ~ rcs(x1, 5) * rcs(x2, 5),
  df = my_df_mice,
  ref_level = "A",
  coef_matrix = TRUE
)