local_approximation: Fit local model around the observation: shortcut for DALEX...

In live: Local Interpretable (Model-Agnostic) Visual Explanations

Description

Fit local model around the observation: shortcut for DALEX explainer objects

Usage

local_approximation(
  explainer,
  observation,
  target_variable_name,
  n_new_obs,
  local_model = "regr.lm",
  select_variables = F,
  predict_type = "response",
  kernel_type = gaussian_kernel,
  ...
)

Arguments

explainer	a model to be explained, preprocessed by the DALEX::explain function
observation	a new observation for which predictions need to be explained
target_variable_name	name of the response variablea as a character
n_new_obs	Number of observation in the simulated dataset
local_model	Character specyfing mlr learner to be used as a local model
select_variables	If TRUE, variable selection will be performed while fitting the local linear model
predict_type	Argument passed to mlr::makeLearner() argument "predict.type" while fitting the local model. Defaults to "response"
kernel_type	Function which will be used to calculate distances from simulated observation to explained instance
...	Arguments to be passed to sample_locally function

Value

object of class live_explainer. More details in fit_explanation function help.

Examples

## Not run: 
data('wine')
library(randomForest)
library(DALEX)
rf <- randomForest(quality~., data = wine)
expl <- explain(rf, wine, wine$quality)
live_expl <- local_approximation(expl, wine[5, ], "quality", 500)

## End(Not run)

live documentation built on Jan. 17, 2020, 9:06 a.m.