predict_tweaked {glmtree} | R Documentation |
Tweaked predict to take into account levels that are not in the training set but in the test set and remove these rows.
Description
This function applies the logistic regression predict after carefully removing observations of categorical features' levels absent from the training set..
Usage
predict_tweaked(model, df, c_iter)
Arguments
model |
The logistic regression model to use to predict on a test set. |
df |
The whole test set which class has to be predicted. |
c_iter |
The segment considered (given by the tree). |
Value
A dataframe of predictions (in rows: the observations, in cols: the class probabilities) given by the model given in input, eventually tweaked if some levels are unknown to it.
Author(s)
Adrien Ehrhardt
[Package glmtree version 0.3.1 Index]