Plots global importance (i.e. cumulative importance of all features for all response variables) and individual response importance.

interpret_Mrvi(
  VI,
  modelPerf,
  Y,
  X,
  groupCov = NULL,
  cutoff = 0.2,
  mode = "regression"
)

Arguments

VI

A dataframe data frame generated from mrvip function

modelPerf

A dataframe data frame generated by modelPerf function

Y

A dataframe response data set

X

A dataframe feature data set

cutoff

A numeric determines Mathews correlation coefficient (mcc) threshold for displaying individual reponse model. Default is 0.5

mode

character'classification' or 'regression' i.e., is the generative model a regression or classification?

Details

1st plot: and individual response feature plots (2nd plot). Requires object generated from the R function mrvip as well as feature data (Y). Variables also need to be grouped to allow for easier interpretation.