gbm.factorplot function

Creates ggplots of marginal effect for factorial variables from plot.gbm in gbm.auto.

Creates an additional plot to those created by gbm.plot within gbm.auto. Can also take Bin/Gaus_Best_line.csv or similar csvs directly. Allows changing of x axis levels and all ggplot and ggsave params.

gbm.factorplot(
  x,
  factorplotlevels = NULL,
  ggplot2guideaxisangle = 0,
  ggplot2labsx = "",
  ggplot2labsy = "Marginal Effect",
  ggplot2axistext = 1.5,
  ggplot2axistitle = 2,
  ggplot2legendtext = 1,
  ggplot2legendtitle = 1.5,
  ggplot2legendtitlealign = 0,
  ggplot2plotbackgroundfill = "white",
  ggplot2plotbackgroundcolour = "grey50",
  ggplot2striptextx = 2,
  ggplot2panelbordercolour = "black",
  ggplot2panelborderfill = NA,
  ggplot2panelborderlinewidth = 1,
  ggplot2legendspacingx = grid::unit(0, "cm"),
  ggplot2legendbackground = ggplot2::element_blank(),
  ggplot2panelbackgroundfill = "white",
  ggplot2panelbackgroundcolour = "grey50",
  ggplot2panelgridcolour = "grey90",
  ggplot2legendkey = ggplot2::element_blank(),
  ggsavefilename = paste0(lubridate::today(), "_Categorical-variable.png"),
  ggsaveplot = ggplot2::last_plot(),
  ggsavedevice = "png",
  ggsavepath = "",
  ggsavescale = 2,
  ggsavewidth = 10,
  ggsaveheight = 4,
  ggsaveunits = "in",
  ggsavedpi = 300,
  ggsavelimitsize = TRUE,
  ...
)

Arguments

• x: Input data.frame or tibble or csv (full file address including .csv) to read, must be a categorical variable.
• factorplotlevels: Character vector of the variable's levels to reorder the x axis by, all must match those in the first column of the csv exactly. Default NULL orders from high to low Y value.
• ggplot2guideaxisangle: Default 0. Set at e.g. 90 to rotate.
• ggplot2labsx: Default: "".
• ggplot2labsy: Default: "Marginal Effect".
• ggplot2axistext: Default: 1.5.
• ggplot2axistitle: Default: 2.
• ggplot2legendtext: Default: 1.
• ggplot2legendtitle: Default: 1.5.
• ggplot2legendtitlealign: Default: 0, # otherwise effect type title centre aligned for some reason.
• ggplot2plotbackgroundfill: Default: "white", white background.
• ggplot2plotbackgroundcolour: Default: "grey50", background lines.
• ggplot2striptextx: Default: 2.
• ggplot2panelbordercolour: Default: "black".
• ggplot2panelborderfill: Default: NA.
• ggplot2panelborderlinewidth: Default: 1.
• ggplot2legendspacingx: Default: unit(0, "cm"), # compress spacing between legend items, this is min.
• ggplot2legendbackground: Default: ggplot2::element_blank().
• ggplot2panelbackgroundfill: Default: "white".
• ggplot2panelbackgroundcolour: Default: "grey50".
• ggplot2panelgridcolour: Default: "grey90".
• ggplot2legendkey: Default: ggplot2::element_blank().
• ggsavefilename: Default: paste0(saveloc, lubridate::today(), "_SankeyAlluvial_EMT.SoEv-EfTyp_Col-EfSz.png").
• ggsaveplot: Default: last_plot().
• ggsavedevice: Default: "png".
• ggsavepath: Default: "".
• ggsavescale: Default: 2.
• ggsavewidth: Default: 10.
• ggsaveheight: Default: 4.
• ggsaveunits: Default: "in".
• ggsavedpi: Default: 300.
• ggsavelimitsize: Default: TRUE.
• ...: Allow params to be called from higher function esp gbm.auto.

Returns

Factorial ggplot saved with users preferred location and name.

Details

`r lifecycle::badge("experimental")

Author(s)

Simon Dedman, simondedman@gmail.com