interaction_heatmap: Plot Variable Interaction Heatmap

In peterbat1/fitMaxnet: Fit MaxEnt Niche Models Using maxnet

Plot Variable Interaction Heatmap

Description

Plot Variable Interaction Heatmap

Usage

interaction_heatmap(
  feature_summary_table = "",
  num_repl = NULL,
  factor_ordering = NULL,
  min_colour = "white",
  max_colour = "blue"
)

Arguments

feature_summary_table	Data.frame. Data table produced by function feature_summary
num_repl	Integer. Number of replicate models used to prepare feature_summary_table
factor_ordering	Character array. Variable names in an order preferred for plotting. Default of NULL uses the
min_colour	Colour name or hexadecimal colour value. Colour for the start of the colour gradient used to shade heatmap grid cells. Default is "white"
max_colour	Colour name or hexadecimal colour value. Colour for the upper level of the colour gradient. Default is "blue"

Details

The data provided in the parameter feature_summary_table is the number of replicate models in which each feature was retained (ie was given a non-zero coefficient).

This does not represent feature importance. Some indirect inference about this may come from considering the importance of each of the two variables contributing to a feature as reported by varImportance.

It might be possible to compute an estimate of feature importance using the same approach implemented in varImportance.

This function is based on code found here: https://r-graph-gallery.com/79-levelplot-with-ggplot2.html

Value

A ggplot2 object

peterbat1/fitMaxnet documentation built on Sept. 17, 2024, 10:50 p.m.