plotHeat: plotHeat

In AlanInglis/vividOld: Variable Importance and Variable Interaction Displays

Description

Plots a Heatmap-tyle display showingVariable Importance and Variable Interaction

Usage

plotHeat(
  dinteraction,
  plotly = FALSE,
  top = NULL,
  title = "",
  intPal = rev(sequential_hcl(palette = "Blues 3", n = 11)),
  impPal = rev(sequential_hcl(palette = "Reds 3", n = 11)),
  fitlimsInt = NULL,
  fitlimsImp = NULL,
  angle = NULL,
  ...
)

Arguments

plotly	If TRUE then an interactive plot is displayed.
top	Returns the first part of the interaction matrix and resulting plot. Similar to head() function.
title	Adds title to the plot.
intPal	A colorspace colour palette to display the interaction values.
impPal	A colorspace colour palette to display the importance values.
fitlimsInt	Specifies the fit range for the color map for interaction strength.
fitlimsImp	Specifies the fit range for the color map for importance.
angle	The angle to display the x-axis labels.
...	Not currently implemented.
mat	A matrix of values to be plotted. Either added by the user or created using the prepFunc() function.
minInt	Minimum interaction strength to be displayed on the legend.
maxInt	Maximum interaction strength to be displayed on the legend.
minImp	Minimum importance value to be displayed on the legend.
maxImp	Maximum importance value to be displayed on the legend.

Examples

# Load in the data:
aq <- data.frame(airquality)
aq <- na.omit(aq)

# Run an mlr ranger model:
library(mlr3)
library(mlr3learners)
library(ranger)
aq <- na.omit(airquality)
aq_Task <- TaskRegr$new(id = "airQ", backend = aq, target = "Ozone")
aq_lrn <- lrn("regr.ranger", importance = "permutation")
aq_Mod <- aq_lrn$train(aq_Task)

# Create matrix
myMat <- vividMatrix(task = aq_Task,  model = aq_Mod)

# Create plot:
plot(myMat, type = "heatMap")

AlanInglis/vividOld documentation built on March 4, 2021, 12:44 a.m.