DALEXtra (version 1.3.2)

yhat.WrappedModel: Wrapper over the predict function

Description

These functios are default predict functions. Each function returns a single numeric score for each new observation. Those functions are very important since informations from many models have to be extracted with various techniques.

Usage

# S3 method for WrappedModel
yhat(X.model, newdata, ...)
# S3 method for H2ORegressionModel
yhat(X.model, newdata, ...)
# S3 method for H2OBinomialModel
yhat(X.model, newdata, ...)
# S3 method for H2OMultinomialModel
yhat(X.model, newdata, ...)
# S3 method for scikitlearn_model
yhat(X.model, newdata, ...)
# S3 method for keras
yhat(X.model, newdata, ...)
# S3 method for LearnerRegr
yhat(X.model, newdata, ...)
# S3 method for LearnerClassif
yhat(X.model, newdata, ...)
# S3 method for GraphLearner
yhat(X.model, newdata, ...)
# S3 method for xgb.Booster
yhat(X.model, newdata, ...)

Arguments

X.model

object - a model to be explained

newdata

data.frame or matrix - observations for prediction

...

other parameters that will be passed to the predict function

Value

An numeric vector of predictions

Details

Currently supported packages are:

• mlr see more in explain_mlr

• h2o see more in explain_h2o

• scikit-learn see more in explain_scikitlearn

• keras see more in explain_keras

• mlr3 see more in explain_mlr3

• xgboost see more in explain_xgboost