mbrPlotInteraction: Plot interaction effects between two continuous variables.
In RTNduals: Analysis of co-regulation and inference of 'dual regulons'

Description Usage Arguments Value Examples

View source: R/AllPlotsMBR.R

This function plots the interaction effects between two continuous variables for linear, logistic, or Cox models.

mbrPlotInteraction(
  model,
  vars,
  xlim = NULL,
  ylim = NULL,
  zlim = NULL,
  xlab = NULL,
  ylab = NULL,
  zlab = NULL,
  zcenter = NULL,
  zlog = NULL,
  zcols = c("#008080ff", "#d45500ff"),
  ycols = c("#4A97C9", "#D92522"),
  showdata = FALSE,
  datacols = "grey50",
  fname = "interplot",
  fpath = ".",
  width = 4.5,
  height = 4,
  plotype = c("3D", "2D"),
  plotpdf = FALSE
)

`model`	An object of class 'lm', 'glm', or 'coxph'.
`vars`	A character vector of length 2 with the names of two varibles in the 'model'.
`xlim`	A numeric vector of length 2, i.e. xlim = c(x1, x2), indicating the x limits of the plot. If xlim = NULL, it will be derevided from the observed data ranges.
`ylim`	A numeric vector of length 2, i.e. ylim = c(y1, y2), indicating the y limits of the plot. If ylim = NULL, it will be derevided from the observed data ranges.
`zlim`	A numeric vector of length 2, i.e. zlim = c(z1, z2), indicating the z limits of the plot. If zlim = NULL, it will be derevided from the observed data ranges.
`xlab`	A string with the label for the x-axis.
`ylab`	A string with the label for the y-axis.
`zlab`	A string with the label for the z-axis.
`zcenter`	A numeric value indicating a z value to center the color scale.
`zlog`	A logical value indicating whether the z axis is to be logarithmic.
`zcols`	A vector of length 2 indicating a diverging color scheme for the z-axis varible.
`ycols`	A vector of length 2 indicating a diverging color scheme for the y-axis varible (only used when type='2D').
`showdata`	A logical value indicating whether to show the original data used to fit linear model.
`datacols`	When 'showdata = TRUE', this can be a named vector of colors for data points (names should match samples in the 'model' object). Alternativelly, it can be a single color value.
`fname`	A string. The name of the PDF file which will contain the plot.
`fpath`	A string. The directory where the file will be saved.
`width`	A numeric value. The width of the plot.
`height`	A numeric value. The height of the plot.
`plotype`	A string indicating '2D' of '3D' plot type. If plotype = '2D', the z-axis (and all related parameters) is transposed to the y-axis.
`plotpdf`	A logical value.

A interaction plot.

#-- Example of simulated data, with response variable modelled by:
#-- (1) Main effects of 'reg1' and 'reg2'
#-- (2) Interaction effects between 'reg1' and 'reg2' 
#-- (3) Additional random uniform noise
reg1 <- rnorm(1000,0,2)
reg2 <- rnorm(1000,0,2)
response <- 3*reg1 + 2*reg1*reg2 + runif(1000,0,2)
dataset <- data.frame(reg1, reg2, response)
model <- lm(response ~ reg1*reg2, data=dataset)
mbrPlotInteraction(model, vars=c("reg1","reg2"))