ggMultiInfluence: ggMultiInfluence

In JBjouffray/ggBRT: Explore and Visualise the Results of Boosted Regression Trees

ggMultiInfluence

Description

Function to return and plot (heatmap) the relative influence of predictors from several boosted regression trees obtained with the gbm.step routine in the dismo package. The models need to have the same set of predictors.

Usage

ggMultiInfluence(..., col.gradient = c("white", "lightblue4"), round = 1,
  col.text = "grey10", col.grid = NULL, size.grid = 0.3,
  legend.pos = "bottom", legend.dir = "horizontal", scale.gradient = c(0,
  100))

Arguments

...	several gbm.step objects (object of S3 class gbm)
col.gradient	a vector of two colors to define the color gradient of the plot
round	indicates the number of decimal places to be used to round the relative influence values (default = 1)
col.text	color of the text within the grid cells
col.grid	color of the contours of the cells (default= NULL)
size.grid	thickness of the contours of the cells (if col.grid is provided)
legend.pos	position for the legend ("none", "right", "left", "bottom" (default), "top")
legend.dir:	direction for the legend ("vertical", "horizontal" (default))
scale.gradient:	a vector with min and max limits for the gradient values (default=c(0,100))

Value

Returns a dataframe with the relative influence of each predictor for the different models and plot a heatmap

JBjouffray/ggBRT documentation built on Nov. 19, 2024, 2:34 a.m.