Requires the future.apply package

find_permuted_auc(
  method,
  test_data,
  outcome_colname,
  feat,
  outcome_value,
  seed
)

Arguments

method

ML method. Options: `c("regLogistic", "rf", "rpart2", "svmRadial", "xgbTree")``

test_data

held out test data: dataframe of outcome and features

outcome_colname

Column name as a string of the outcome variable (default NULL; will be chosen automatically).

feat

feature or group of correlated features to permute

outcome_value

Outcome value of interest as a string (default NULL; will be chosen automatically).

seed

Random seed (default: NA). Your results will be reproducible if you set a seed.

Value

vector of mean permuted auc and mean difference between test and permuted auc

Author

Begüm Topçuoğlu, topcuoglu.begum@gmail.com

Zena Lapp, zenalapp@umich.edu