explain: Create Model Explainer

Description Usage Arguments Details Value Examples

Description

Black-box models may have very different structures. This function creates a unified representation of a model, which can be further processed by various explainers.

Usage

1
2
3
4
5
6
explain.default(model, data = NULL, y = NULL,
  predict_function = yhat, link = I, ..., label = tail(class(model),
  1))

explain(model, data = NULL, y = NULL, predict_function = yhat,
  link = I, ..., label = tail(class(model), 1))

Arguments

model

object - a model to be explained

data

data.frame or matrix - data that was used for fitting. If not provided then will be extracted from the model

y

numeric vector with outputs / scores. Currently used only by variable_dropout() explainer.

predict_function

function that takes two arguments: model and new data and returns numeric vector with predictions

link

function - a transformation/link function that shall be applied to raw model predictions

...

other parameters

label

character - the name of the model. By default it's extracted from the 'class' attribute of the model

Details

Please NOTE, that the model is actually the only required argument. But some explainers may require that others will be provided too.

Value

An object of the class 'explainer'.

It's a list with following fields:

Examples

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
apartments_lm <- lm(m2.price ~ ., data = apartments)
apartments_lm_ex <- explain(apartments_lm, data = apartments, label = "apartments_lm")
apartments_lm_ex

 ## Not run: 
library("breakDown2")
wine_lm_model4 <- lm(quality ~ pH + residual.sugar + sulphates + alcohol, data = wine)
wine_lm_explainer4 <- explain(wine_lm_model4, data = wine, label = "model_4v")
wine_lm_explainer4

library("randomForest")
wine_rf_model4 <- randomForest(quality ~ pH + residual.sugar + sulphates + alcohol, data = wine)
wine_rf_explainer4 <- explain(wine_rf_model4, data = wine, label = "model_rf")
wine_rf_explainer4
 
## End(Not run)

pbiecek/DALEX2 documentation built on Dec. 7, 2018, 7:55 a.m.