Visualize the Joint Effect of Variables in a GLM

This function visualizes the joint effect of several variables in a generalized linear model (GLM) by plotting the predicted response over a range of values for the primary predictor variable, while holding other predictors at their median or mode values. It allows specifying a group for focused analysis.

visglm(
  demogmdl,
  qmdl,
  x,
  y,
  group = NULL,
  titlestring = "",
  varstoadd = NULL,
  groupvar = "group",
  predictorsigns = NULL,
  jdf_simulation = FALSE,
  xrange = NULL,
  verbose = FALSE
)

Arguments

demogmdl: Data frame containing the variables used in the GLM.
qmdl: Fitted GLM object from which predictions will be generated.
x: Character vector specifying the names of the predictor variables, with the first being the primary.
y: The name of the response variable.
group: The name of the group variable or specific group to be analyzed.
titlestring: Title of the plot indicating the focus of the visualization.
varstoadd: optional additional variables to add into the model; otherwise just extract from the equation
groupvar: Optional; the name of the variable in demogmdl that defines group membership. Default is 'group'.
predictorsigns: Optional; a named numeric vector indicating the direction of the effect of each predictor.
jdf_simulation: boolean
xrange: explicitly set xrange
verbose: Logical; if TRUE, additional processing information will be printed to the console.

Value

Generates a plot visualizing the predicted response and confidence intervals across the range of the primary predictor.

Examples

# Assuming `data` is your dataset, `fit` is a fitted GLM, 
# and you're interested in predictors `x1` and `x2`:
# visglm(data, fit, c("x1", "x2"), "y", "control", 
#   "Visualization for Control Group")
#