Area Under the Curve of Binomial GLM predictions vs. observations

Fits a Quasibinomial GLM model y ~ x with the binomial response y (values 0 and 1) and the numeric, character, or factor predictor x using stats::glm() and returns the area under the ROC curve of the observations against the predictions (see score_auc()).

Cases are weighted with case_weights() to prevent issues arising from class imbalance.

Supports cross-validation via the arguments arguments cv_training_fraction (numeric between 0 and 1) and cv_iterations (integer between 1 and n) introduced via ellipsis (...). See preference_order() for further details.

Usage

f_binomial_glm(df, ...)

Arguments

df

(required, dataframe) with columns:

• "x": (numeric, character, factor) predictor.

• "y" (integer) binomial response with unique values 0 and 1.

...

(optional) Accepts the arguments cv_training_fraction (numeric between 0 and 1) and cv_iterations (integer between 1 and Inf) for cross validation.

Value

numeric or numeric vector: AUC

See also

Other preference_order_functions: f_binomial_gam(), f_binomial_rf(), f_categorical_rf(), f_count_gam(), f_count_glm(), f_count_rf(), f_numeric_gam(), f_numeric_glm(), f_numeric_rf(), preference_order()

Examples

data(vi_smol)

df <- data.frame(
  y = vi_smol[["vi_binomial"]],
  x = vi_smol[["swi_max"]]
)

#no cross-validation
f_binomial_glm(df = df)
#> [1] 0.7161608

#cross-validation
f_binomial_glm(
  df = df,
  cv_training_fraction = 0.5,
  cv_iterations = 10
  )
#>  [1] 0.7342075 0.7601301 0.7172146 0.7097408 0.7183745 0.7152365 0.6981429
#>  [8] 0.7284071 0.7129516 0.6773125

#categorical predictor
df <- data.frame(
  y = vi_smol[["vi_binomial"]],
  x = vi_smol[["koppen_zone"]]
)

f_binomial_glm(df = df)
#> [1] 0.9334159