EffectsPlotter: Visualize model effects

In huntercole25/EcoCountHelper: An assistant for comparing and interpretting models with count data

Visualize model effects

Description

This function produces effects plots for the conditional model of model objects produced by glmmTMB. Effects can be plotted for multiple models in a single function call. These effects plots are based on those found in this document.

Usage

EffectsPlotter(
  TopMods,
  ParamLabs = NULL,
  ConfInts = c(90, 80),
  Scaled = T,
  ThemeBlack = T
)

Arguments

TopMods	This argument can accept multiple input types, and is used to specify which models effects plots should be generated for. If an effects plot is desired for a single model, users can supply the unquoted model object name. If effects plots for multiple models are desired, users can specify a chahracter vector of model names.
ParamLabs	A character vector of labels for each model estimate. Labels must be supplied for estimates in the same order as the vector returned by row.names(summary(x)$coeff$cond), where x is a model object name.
ConfInts	A numeric vector specifying up to three condifence intervals as percent values from largest to smallest (e.g., for 80%, 90%, and 95% confidence intervals, supply ConfInts = c(95, 90, 80)).
Scaled	A logical value indicating whether model estimates are scaled (T) or not (F). When Scaled = T, there will be a dotted geom_abline where estimates = 0.
ThemeBlack	A logical value indicating whether theme_nocturnal should be applied to plots (T) or not (F).

Value

This function generates effects plots using geom_point for point estimates and geom_crossbar for confidence intervals. Resultant plots are named with the model name followed by "EffectsPlot" (e.g., for a model named "EpfuNb2", the effects plot would be named "EpfuNb2EffectsPlot").

Details

Because all models specified in a call to this function share estimate labels, all models in a a given call to this function must have the same model formula.

Examples

data("Epfu_Nb2_Long", "Myev_Nb2_Long", package = "EcoCountHelper")
Labels <- letters[1:12]

#Effects plot for a single model

EffectsPlotter(Epfu_Nb2_Long, Labels)
Epfu_Nb2_LongEffectsPlot

#Effects plot for multiple models specified
#in a character vector.

Mods <- c("Epfu_Nb2_Long", "Myev_Nb2_Long")
EffectsPlotter(Mods, Labels)
Epfu_Nb2_LongEffectsPlot
Myev_Nb2_LongEffectsPlot

#Effects plot for multiple models specified
#in a data frame

ModTable <- data.frame(Species = c("Epfu", "Myev"),
                        Mods = c("Epfu_Nb2_Long", "Myev_Nb2_Long"))
                        
EffectsPlotter(ModTable$Mods, Labels)

Epfu_Nb2_LongEffectsPlot
Myev_Nb2_LongEffectsPlot

huntercole25/EcoCountHelper documentation built on Jan. 14, 2023, 4:13 a.m.