ggResponse2: Partial dependence plot by ggplot with 95 percent confidence...

In rvalavi/myspatial: A collection of functions for spatial data analysis and modelling

Description

This function can be used to plot a partial dependence plot for any model that can predict on data.frames. This function calculates the 95 percent confidence interval created by multiple models (from cross-validation, bootstrapping or multi-model) around each curve.

Usage

ggResponse2(
  models,
  covariates,
  colPlot = 3,
  responseName = "Prediction",
  index = 2,
  ...
)

Arguments

models	a list of model objects (several fitted models on the same dataset).
covariates	the covariates used in model fitting, a raster or data.frame
colPlot	integer. The number of colums for plotting
responseName	character. the name for y axes
index	integer. The columns used for prediction. This relates to the factor level for classification.
...	other arguments e.g. type = 'response' in GLMs or type = 'prob' in randomForest or the number of trees in BRT/GBM.

Author(s)

Roozbeh Valavi

rvalavi/myspatial documentation built on May 15, 2021, 12:14 a.m.