midr (version 0.5.0)

get.yhat: Wrapper Prediction Function

Description

get.yhat() works as a proxy prediction function for many classes of fitted models.

Usage

get.yhat(X.model, newdata, ...)
# S3 method for default
get.yhat(X.model, newdata, target = -1L, ...)
# S3 method for mid
get.yhat(X.model, newdata, ...)
# S3 method for lm
get.yhat(X.model, newdata, ...)
# S3 method for glm
get.yhat(X.model, newdata, ...)
# S3 method for rpart
get.yhat(X.model, newdata, target = -1L, ...)
# S3 method for randomForest
get.yhat(X.model, newdata, target = -1L, ...)
# S3 method for ranger
get.yhat(X.model, newdata, target = -1L, ...)
# S3 method for svm
get.yhat(X.model, newdata, target = -1L, ...)
# S3 method for ksvm
get.yhat(X.model, newdata, target = -1L, ...)
# S3 method for AccurateGLM
get.yhat(X.model, newdata, ...)
# S3 method for glmnet
get.yhat(X.model, newdata, ...)
# S3 method for model_fit
get.yhat(X.model, newdata, target = -1L, ...)
# S3 method for rpf
get.yhat(X.model, newdata, target = -1L, ...)

Value

get.yhat() returns a numeric vector of model predictions for the newdata.

Arguments

X.model

a fitted model object.

newdata

a data.frame or matrix.

...

optional parameters that are passed to the prediction method for the model.

target

an integer or character vector specifying the target levels for the prediction, used for the models that returns a matrix or data.frame of class probabilities. Default is -1, representing the probability of not being the base level.

Details

get.yhat() is a wrapper prediction function for many classes of models. Although many predictive models have their own method of stats::predict(), the structure and the type of the output of these methods are not uniform. get.yhat() is designed to always return a simple numeric vector of model predictions. The design of get.yhat() is strongly influenced by DALEX::yhat().

Examples

data(trees, package = "datasets")
model <- glm(Volume ~ ., trees, family = Gamma(log))
predict(model, trees[1:5, ], type = "response")
get.yhat(model, trees[1:5, ])