Interpret a model via regularized coefficient estimates

Interpret a model via regularized coefficient estimates

interpret(x, sparsity = NULL, remove_zeros = TRUE, top_n)

Arguments

x	a model_list object containing a glmnet model
sparsity	If NULL (default) coefficients for the best-performing model will be returned. Otherwise, a value in [0, 1] that determines the sparseness of the model for which coefficients will be returned, with 0 being maximally sparse (i.e. having the fewest non-zero coefficients) and 1 being minimally sparse
remove_zeros	Remove features with coefficients equal to 0? Default is TRUE
top_n	Integer: How many coefficients to return? The largest top_n absolute-value coefficients will be returned. If missing (default), all coefficients are returned

Value

A data frame of variables and their regularized regression coefficient estimates with parent class "interpret"

Details

**WARNING** Coefficients are on the scale of the predictors; they are not standardized, so unless features were scaled before training (e.g. with prep_data(..., scale = TRUE), the magnitude of coefficients does not necessarily reflect their importance.

If x was trained with more than one value of alpha the best value of alpha is used; sparsity is determined only via the selection of lambda. Using only lasso regression (i.e. alpha = 1) will produce a sparser set of coefficients and can be obtained by not tuning hyperparameters.

See also

plot.interpret

Examples

m <- machine_learn(pima_diabetes, patient_id, outcome = diabetes, models = "glm")
#> Training new data prep recipe...
#> Variable(s) ignored in prep_data won't be used to tune models: patient_id
#> 
#> diabetes looks categorical, so training classification algorithms.
#> 
#> After data processing, models are being trained on 12 features with 768 observations.
#> Based on n_folds = 5 and hyperparameter settings, the following number of models will be trained: 100 glm's 
#> Training with cross validation: glmnet
#> 
#> *** Models successfully trained. The model object contains the training data minus ignored ID columns. ***
#> *** If there was PHI in training data, normal PHI protocols apply to the model object. ***
interpret(m)
#> Reference Levels:
#> All `weight_class` estimates are relative to `obese`
#> 
#> # A tibble: 9 x 2
#>   variable                                coefficient
#> * <chr>                                         <dbl>
#> 1 (Intercept)                                -5.44   
#> 2 weight_class_normal (vs. obese)            -1.35   
#> 3 weight_class_overweight (vs. obese)        -0.554  
#> 4 pedigree                                    0.553  
#> 5 weight_class_morbidly.obese (vs. obese)     0.118  
#> 6 pregnancies                                 0.0866 
#> 7 plasma_glucose                              0.0322 
#> 8 skinfold                                    0.00660
#> 9 age                                         0.00643
interpret(m, .2)
#> Reference Levels:
#> All `weight_class` estimates are relative to `obese`
#> 
#> # A tibble: 4 x 2
#>   variable                        coefficient
#> * <chr>                                 <dbl>
#> 1 (Intercept)                         -3.75  
#> 2 weight_class_normal (vs. obese)     -0.208 
#> 3 plasma_glucose                       0.0247
#> 4 pregnancies                          0.0197
interpret(m) %>%
  plot()