Skip to contents

Spatial LIME

Jakub Nowosad

2025-01-27

Source: vignettes/articles/Spatial-LIME.Rmd
Spatial-LIME.Rmd

The spatialexplain package provides model agnostic tools for exploring and explaining spatial machine learning models. This vignette shows a simple example of how to use it to explain a regression model with a few implementations of the Local Interpretable Model-agnostic Explanations (LIME) method.

Let’s start by attaching the necessary packages, reading the predictors raster, and loading the pre-trained regression model.

# attaching the necessary packages
library(spatialexplain)
library(terra)

# reading the predictors raster
predictors = rast("/vsicurl/https://github.com/Nowosad/IIIRqueR_workshop_materials/raw/refs/heads/main/data/predictors.tif")
plot(predictors, axes = FALSE)


# loading the pre-trained regression model
data("regr_exp", package = "spatialexplain")
regr_exp
#> Model label:  rpart 
#> Model class:  rpart 
#> Data head  :
#>    popdens    coast      dem      ndvi  lst_day lst_night
#> 1 0.000000 1.126301 85.90540 0.3656146 24.37792  12.64256
#> 2 1.211701 6.743273 75.00126 0.3990190 28.13341  10.70668

The regr_exp object is a pre-trained regression model that was trained on the predictors raster.

The main idea behind the LIME method is to approximate the complex model with a simpler one that is easier to interpret. The map_lime() function explains the model using one of three implementations of the LIME method: type = "localModel", type = "iml", and type = "lime".

The default one is the "localModel" method.

regr_lime1 = map_lime(regr_exp, predictors, maxcell = 500, type = "localModel")
plot(regr_lime1)

The second one is the "iml" method1.

regr_lime2 = map_lime(regr_exp, predictors, maxcell = 500, type = "iml")
#> Warning in private$aggregate(): Had to choose a smaller k
plot(regr_lime2)

The last one is the "lime" method; however, currently, it only works for explainers created with one of the supported models/modeling framework (e.g., ranger or caret).

regr_lime3 = map_lime(regr_exp, predictors, maxcell = 500, type = "lime")

These different implementations use other algorithms to extract the interpretable features, have other sampling methods, and other ways of weighting. For more explanation of the LIME method, read the “Local Interpretable Model-agnostic Explanations (LIME)” chapter from the Explanatory Model Analysis book.