plot_predictions: Plot model predictions for 'pffr' models

In bauer-alex/FoSIntro: Convenience Functions for Semiparametric Function-on-Scalar Regression

Description

Plots the model predictions for a function-on-scalar model fitted with pffr.

Usage

plot_predictions(model, newdata, yvar, ci_type = "none",
  ci_alpha = 0.05, labels = NULL, legend_title = "group",
  col_palette = "Set1", col_vector = NULL, rev_cols = FALSE,
  base_size = 11, legend_symbol_size = 2, type = "response",
  log10 = FALSE, plot.margin = unit(rep(5.5, 4), "points"), hline,
  ybreaks = waiver(), ylabels = waiver(), xlab = "yindex",
  ylab = "prediction", main, ylim = NULL, lwd = 1, ...)

Arguments

model	Function-on-scalar model fitted with pffr
newdata	See predict.pffr
yvar	name of the response variable in model
ci_type	One of "none" (default; no confidence bands), "ci" (pointwise confidence intervals for the predicted mean) or "pi_norm"/"pi_gamma" (pointwise prediction intervals), specifying which type of uncertainty bands should be plotted. Prediction intervals are either calculated summing the predicted mean variance and the residual variance ("pi_norm") or using 'posterior simulation' of the parameter vector ("pi_gamma") as outlined here: https://stat.ethz.ch/pipermail/r-help/2011-April/275632.html. Note that the "pi_norm" PIs are not generally applicable for generalized models. The posterior simulation approach is more generally applicable, but is currently only implemented for Gamma regression.
ci_alpha	Alpha to be used for confidence or prediction intervals. Defaults to 0.05, i.e. 95% intervals.
labels	Labels vector with length equal to nrow(newdata). Used for the legend.
legend_title	Title of legend
col_palette	Color palette to use in plot, see scale_color_brewer. Only used if nrow(newdata)>1 and if is.null(col_vector).
col_vector	Vector of colors to use
rev_cols	If TRUE, the color palette is turned around.
base_size	Base size of plot elements, see theme_bw. Defaults to 11.
legend_symbol_size	Size of symbols in legend. Default to 2
type	type argument for predict.pffr
log10	If TRUE the predictions are shown on a log10 scale
plot.margin	Margin of the ggplot
hline	Vector of geom_hline
ybreaks	Numeric vector of breaks for the y axis
ylabels	Character vector of labels for the y axis. Only used if ybreaks is specified
xlab, ylab, main, ylim, lwd	Further arguments for the plot function
...	Additional arguments passed to pi_gamma

bauer-alex/FoSIntro documentation built on Feb. 11, 2022, 8:33 a.m.