Tidying methods for Spark ML linear models

R/tidiers_ml_linear_models.R

ml_glm_tidiers

Description

These methods summarize the results of Spark ML models into tidy forms.

Usage

## S3 method for class 'ml_model_generalized_linear_regression'
tidy(x, exponentiate = FALSE, ...) 

## S3 method for class 'ml_model_linear_regression'
tidy(x, ...) 

## S3 method for class 'ml_model_generalized_linear_regression'
augment( 
  x, 
  newdata = NULL, 
  type.residuals = c("working", "deviance", "pearson", "response"), 
  ... 
) 

## S3 method for class '`_ml_model_linear_regression`'
augment( 
  x, 
  new_data = NULL, 
  type.residuals = c("working", "deviance", "pearson", "response"), 
  ... 
) 

## S3 method for class 'ml_model_linear_regression'
augment( 
  x, 
  newdata = NULL, 
  type.residuals = c("working", "deviance", "pearson", "response"), 
  ... 
) 

## S3 method for class 'ml_model_generalized_linear_regression'
glance(x, ...) 

## S3 method for class 'ml_model_linear_regression'
glance(x, ...) 

Arguments

Arguments Description
x a Spark ML model.
exponentiate For GLM, whether to exponentiate the coefficient estimates (typical for logistic regression.)
extra arguments (not used.)
newdata a tbl_spark of new data to use for prediction.
type.residuals type of residuals, defaults to "working". Must be set to "working" when newdata is supplied.
new_data a tbl_spark of new data to use for prediction.

Details

The residuals attached by augment are of type “working” by default, which is different from the default of “deviance” for residuals() or sdf_residuals().