xspline: Builds predictive model based GLM.

In krystian8207/xspliner: Assisted Model Building, using Surrogate Black-Box Models to Train Interpretable Spline Based Additive Models

Description

The method provides main functionality on building GLM models with automatic variables transformation. The transformations are based on specified single variable responses for selected black-box model. See details in vignette("xspliner").

Usage

xspline(object, ...)

model_surrogate_xspliner(object, ...)

## Default S3 method:
xspline(object, lhs = NULL, response = NULL,
  predictors = NULL, data = NULL, form = "additive", bare = NULL,
  env = parent.frame(), ...)

## S3 method for class 'formula'
xspline(object, model, data = NULL,
  consider = "specials", env = parent.frame(), ...)

## S3 method for class 'explainer'
xspline(object, env = parent.frame(), ...)

Arguments

object	Predictive model, formula or explainer (see DALEX) object.
...	Other arguments passed to xspline methods or build_xspliner.
lhs	Left-hand side of model formula. Can be transformed response.
response	Name of response variable of model.
predictors	Predictor values that should be used in final model.
data	Training data of model.
form	Can be 'additive' (default) or 'multiplicative'. Specifies formula form in final model.
bare	Variable names that mustn't be transformed in final model.
env	Environment in which optional variables passed into parameters are stored. variables transformation. See vignette("xspliner") for details.
model	When object is formula - predictive model. Basic model used for extracting predictors transformation.
consider	One of c("specials", "all"). If "specials", only components with xs or xf call are considered in transition.

Details

model_surrogate_xspliner is a wrapper of xspline method to assure consistency with https://github.com/ModelOriented/DrWhy tools

Value

GLM object of class 'xspliner'.

Examples

# preparing blackbox model
library(randomForest)
rf_iris <- randomForest(
  Petal.Width ~  Sepal.Length + Petal.Length + Species,
  data = iris)

# formula based xspliner
xs_iris <- xspline(
  Petal.Width ~ xs(Sepal.Length) + xs(Petal.Length) + xf(Species),
  model = rf_iris)
summary(xs_iris)
plot(xs_iris, "Sepal.Length")

# passing just the model
xs_iris <- xspline(rf_iris)
summary(xs_iris)
plot(xs_iris, "Sepal.Length")

# using DALEX
library(DALEX)
xs_iris_explainer <- explain(rf_iris)
xs_iris <- xspline(rf_iris)
summary(xs_iris)
plot(xs_iris, "Sepal.Length")

krystian8207/xspliner documentation built on Oct. 6, 2019, 11:02 p.m.